sentence1
stringlengths
52
3.87M
sentence2
stringlengths
1
47.2k
label
stringclasses
1 value
def __rm_duplicate_citation_2(self, key): """ Remove exact duplicate entries for abstract or citation fields. Check all publication entries. :return: """ citations = [] # Before writing, remove any duplicate "Full_Citations" that are found for idx, pub in enumerate(self.noaa_data_sorted["Publication"]): try: citations.append(pub[key]) except KeyError: # Key was not found. Enter a blank entry in citations list citations.append("") # Create a backwards dictionary, that lists which indexes are duplicates d = defaultdict(list) for i, item in enumerate(citations): d[item].append(i) d = {k: v for k, v in d.items() if len(v) > 1} # Duplicates indexes are listed like so: # {"full citation info here": [0, 2, 5]} # Loop over duplicate indexes for citation, idxs in d.items(): # do not process any duplicate entries that were initially blanks. if citation: # Loop for [1:], since we want to keep the first citation and remove the rest. for idx in idxs[1:]: # Set citation to blank self.noaa_data_sorted["Publication"][idx][key] = ""
Remove exact duplicate entries for abstract or citation fields. Check all publication entries. :return:
entailment
def __rm_names_on_csv_cols(d): """ Remove the variableNames from the columns. :param dict d: Named csv data :return list list: One list for names, one list for column data """ _names = [] _data = [] for name, data in d.items(): _names.append(name) _data.append(data) return _names, _data
Remove the variableNames from the columns. :param dict d: Named csv data :return list list: One list for names, one list for column data
entailment
def __get_author_last_name(author): """ Take a string of author(s), get the first author name, then get the first authors last name only. :param str author: Author(s) :return str _author: First author's last name """ _author = "" if isinstance(author, str): try: # example: 'Ahmed, Moinuddin and Anchukaitis, Kevin J and ...' # If this is a list of authors, try to split it and just get the first author. if " and " in author: _author = author.split(" and ")[0] # 'Ahmed, Moinuddin; Anchukaitis, Kevin J; ...' elif ";" in author: _author = author.split(";")[0] except Exception: _author = "" try: # example : 'Ahmed Moinuddin, Anchukaitis Kevin J, ...' _author = author.replace(" ", "") if "," in author: _author = author.split(",")[0] except Exception: _author = "" elif isinstance(author, list): try: # example: [{'name': 'Ahmed, Moinuddin'}, {'name': 'Anchukaitis, Kevin J.'}, ..] # just get the last name of the first author. this could get too long. _author = author[0]["name"].split(",")[0] except Exception as e: _author = "" return _author
Take a string of author(s), get the first author name, then get the first authors last name only. :param str author: Author(s) :return str _author: First author's last name
entailment
def __create_author_investigator_str(self): """ When investigators is empty, try to get authors from the first publication instead. :return str author: Author names """ _author = "" try: for pub in self.noaa_data_sorted["Publication"]: if "author" in pub: if pub["author"]: _author_src = pub["author"] if isinstance(_author_src, str): try: if " and " in _author_src: _author = _author_src.replace(" and ", "; ") elif ";" in _author_src: # If there is a semi-colon, add a space after it, just in case it didn't have one _author = _author_src.replace(";", "; ") break except Exception as e: _author = "" elif isinstance(_author_src, list): try: for _entry in _author_src: _author += _entry["name"].split(",")[0] + ", " except Exception as e: _author = "" except Exception: _author = "" return _author
When investigators is empty, try to get authors from the first publication instead. :return str author: Author names
entailment
def __get_filename(table): """ Get filename from a data table :param dict table: :return str: """ try: filename = table['filename'] except KeyError as e: filename = "" logger_lpd_noaa.warning("get_filename: KeyError: Table missing filename, {}".format(e)) except TypeError: try: filename = table[0]["filename"] except Exception as e: filename = "" logger_lpd_noaa.warning("get_filename: Generic: Unable to get filename from table, {}".format(e)) return filename
Get filename from a data table :param dict table: :return str:
entailment
def __get_doi(pub): """ Get DOI from this ONE publication entry. :param dict pub: Single publication entry :return: """ doi = "" # Doi location: d["pub"][idx]["identifier"][0]["id"] try: doi = pub["DOI"][0]["id"] doi = clean_doi(doi) except KeyError: logger_lpd_noaa.info("get_dois: KeyError: missing a doi key") except Exception: logger_lpd_noaa.info("get_dois: Exception: something went wrong") # if we received a doi that's a list, we want to concat into a single string if isinstance(doi, list): if len(doi) == 1: doi = doi[0] else: ", ".join(doi) return doi
Get DOI from this ONE publication entry. :param dict pub: Single publication entry :return:
entailment
def __get_noaa_key_w_context(header, lipd_key): """ Use the section header and key to get the proper noaa key. :param str header: :param str lipd_key: :return: """ try: noaa_key = LIPD_NOAA_MAP_BY_SECTION[header][lipd_key] except KeyError: return lipd_key return noaa_key
Use the section header and key to get the proper noaa key. :param str header: :param str lipd_key: :return:
entailment
def __get_overall_data(self, x): """ (recursive) Collect all "sensorGenus" and "sensorSpecies" fields, set data to self :param any x: Any data type :return none: """ if isinstance(x, dict): if "sensorGenus" in x: if x["sensorGenus"] and x["sensorGenus"] not in self.lsts_tmp["genus"]: self.lsts_tmp["genus"].append(x["sensorGenus"]) if "sensorSpecies" in x: if x["sensorSpecies"] and x["sensorSpecies"] not in self.lsts_tmp["species"]: self.lsts_tmp["species"].append(x["sensorSpecies"]) if "archiveType" in x: if x["archiveType"] and x["archiveType"] not in self.lsts_tmp["archive"]: self.lsts_tmp["archive"].append(x["archiveType"]) if "QCnotes" in x: if x["QCnotes"] and x["QCnotes"] not in self.lsts_tmp["qc"]: self.lsts_tmp["qc"].append(x["QCnotes"]) for k, v in x.items(): if isinstance(v, dict): self.__get_overall_data(v) elif isinstance(v, list): self.__get_overall_data(v) elif isinstance(x, list): for i in x: self.__get_overall_data(i) return x
(recursive) Collect all "sensorGenus" and "sensorSpecies" fields, set data to self :param any x: Any data type :return none:
entailment
def __get_max_min_time_1(self, table): """ Search for either Age or Year to calculate the max, min, and time unit for this table/file. Preference: Look for Age first, and then Year second (if needed) :param dict table: Table data :return dict: Max, min, and time unit """ try: # find the values and units we need to calculate vals, units = self.__get_max_min_time_2(table, ["age", "yearbp", "yrbp"], True) if not vals and not units: vals, units = self.__get_max_min_time_2(table, ["year", "yr"], True) if not vals and not units: vals, units = self.__get_max_min_time_2(table, ["age", "yearbp", "yrbp"], False) if not vals and not units: vals, units = self.__get_max_min_time_2(table, ["year", "yr"], False) # now put this data into the noaa data sorted self for writing to file. # year farthest in the past _max = max(vals) _min = min(vals) if _min or _min in [0, 0.0]: self.noaa_data_sorted["Data_Collection"]["Earliest_Year"] = str(_min) if _max or _max in [0, 0.0]: self.noaa_data_sorted["Data_Collection"]["Most_Recent_Year"] = str(_max) # AD or... yrs BP? self.noaa_data_sorted["Data_Collection"]["Time_Unit"] = units except Exception as e: logger_lpd_noaa.debug("get_max_min_time_2: {}".format(e)) return
Search for either Age or Year to calculate the max, min, and time unit for this table/file. Preference: Look for Age first, and then Year second (if needed) :param dict table: Table data :return dict: Max, min, and time unit
entailment
def __get_max_min_time_2(table, terms, exact): """ Search for either Age or Year to calculate the max, min, and time unit for this table/file. Preference: Look for Age first, and then Year second (if needed) :param dict table: Table data :param list terms: age, yearbp, yrbp, or year, yr :param bool exact: Look for exact key match, or no :return bool: found age or year info """ vals = [] unit = "" try: for k, v in table["columns"].items(): if exact: if k.lower() in terms: try: vals = v["values"] unit = v["units"] break except KeyError: pass elif not exact: for term in terms: if term in k: try: vals = v["values"] unit = v["units"] break except KeyError: pass except Exception as e: logger_lpd_noaa.debug("get_max_min_time_3: {}".format(e)) return vals, unit
Search for either Age or Year to calculate the max, min, and time unit for this table/file. Preference: Look for Age first, and then Year second (if needed) :param dict table: Table data :param list terms: age, yearbp, yrbp, or year, yr :param bool exact: Look for exact key match, or no :return bool: found age or year info
entailment
def __get_data_citation(self, l): """ If originalDataURL / investigators not in root data, check for a dataCitation pub entry. :return: """ # loop once for each publication entry for pub in l: try: # at the moment, these are the only keys of interest inside of dataCitation. Check each. for key in ["url", "investigators"]: if pub["type"] == "dataCitation" and key in pub: noaa_key = self.__get_noaa_key(key) self.data_citation[noaa_key] = pub[key] except KeyError: # no "type" key in pub logger_lpd_noaa.info("lpd_noaa: get_data_citation: KeyError: pub is missing 'type'") return
If originalDataURL / investigators not in root data, check for a dataCitation pub entry. :return:
entailment
def __get_output_filenames(self): """ Set up the output filenames. If more than one file is being written out, we need appended filenames. :return: """ # if there is only one file, use the normal filename with nothing appended. if len(self.noaa_data_sorted["Data"]) == 1: self.output_filenames.append(self.filename_txt) # if there are multiple files that need to be written out, (multiple data table pairs), then append numbers elif len(self.noaa_data_sorted["Data"]) > 1: for i in range(0, self.output_file_ct): tmp_name = "{}-{}.txt".format(self.dsn, i+1) self.output_filenames.append(tmp_name) return
Set up the output filenames. If more than one file is being written out, we need appended filenames. :return:
entailment
def __get_table_pairs(self): """ Use the tables in self.paleos and self.chrons (sorted by idx) to put in self.noaa_data_sorted. :return: """ try: for _idx, _p in enumerate(self.data_paleos): _c = {} try: _c = self.data_chrons[_idx] except IndexError: pass # create entry in self object collection of data tables self.noaa_data_sorted["Data"].append({"paleo": _p, "chron": _c}) except KeyError: logger_lpd_noaa.warning("lpd_noaa: get_meas_table: 0 paleo data tables") self.output_file_ct = len(self.noaa_data_sorted["Data"]) return
Use the tables in self.paleos and self.chrons (sorted by idx) to put in self.noaa_data_sorted. :return:
entailment
def __get_table_count(self): """ Check how many tables are in the json :return int: """ _count = 0 try: keys = ["paleo", "paleoData", "paleoMeasurementTable"] # get the count for how many tables we have. so we know to make appended filenames or not. for pd_name, pd_data in self.lipd_data["paleoData"].items(): for section_name, section_data in pd_data.items(): _count += len(section_data) except Exception: pass if _count > 1: self.append_filenames = True return
Check how many tables are in the json :return int:
entailment
def __put_tables_in_self(self, keys): """ Metadata is sorted-by-name. Use this to put the table data into the object self. :param list keys: Paleo or Chron keys :return none: """ _idx = 1 try: # start adding the filenames to the JSON and adding the tables to the self tables for pd_name, pd_data in self.lipd_data[keys[1]].items(): for table_name, table_data in pd_data[keys[2]].items(): if self.append_filenames: _url = "{}/{}-{}.txt".format(self.wds_url, self.dsn, _idx) _url = re.sub("['{}@!$&*+,;?%#~`\[\]=]", "", _url) self.lipd_data[keys[1]][pd_name][keys[2]][table_name]["WDSPaleoUrl"] = _url table_data["WDSPaleoUrl"] = _url if keys[0] == "paleo": self.data_paleos.append(table_data) else: self.data_chrons.append(table_data) else: _url = re.sub("['{}@!$&*+,;?%#~`\[\]=]", "", self.txt_file_url) self.lipd_data[keys[1]][pd_name][keys[2]][table_name]["WDSPaleoUrl"] = _url table_data["WDSPaleoUrl"] = _url if keys[0] == "paleo": self.data_paleos.append(table_data) else: self.data_chrons.append(table_data) _idx += 1 except Exception: pass return
Metadata is sorted-by-name. Use this to put the table data into the object self. :param list keys: Paleo or Chron keys :return none:
entailment
def __create_file(self): """ Open text file. Write one section at a time. Close text file. Move completed file to dir_root/noaa/ :return none: """ logger_lpd_noaa.info("enter create_file") self.__get_output_filenames() for idx, filename in enumerate(self.output_filenames): try: # self.noaa_txt = open(os.path.join(self.path, filename), "w+") # self.noaa_file_output[filename] = "" # self.noaa_txt = self.noaa_file_output[filename] self.noaa_txt = "" print("writing: {}".format(filename)) logger_lpd_noaa.info("write_file: opened output txt file") except Exception as e: logger_lpd_noaa.error("write_file: failed to open output txt file, {}".format(e)) return self.__get_max_min_time_1(self.noaa_data_sorted["Data"][idx]["paleo"]) self.__check_time_values() self.__write_top(filename) self.__write_generic('Contribution_Date') self.__write_generic('File_Last_Modified_Date') self.__write_generic('Title') self.__write_generic('Investigators') self.__write_generic('Description_Notes_and_Keywords') self.__write_pub() self.__write_funding() self.__write_geo() self.__write_generic('Data_Collection') self.__write_generic('Species') self.__write_data(idx) self.noaa_file_output[filename] = self.noaa_txt # logger_lpd_noaa.info("closed output text file") # reset the max min time unit to none self.max_min_time = {"min": "", "max": "", "time": ""} # shutil.copy(os.path.join(os.getcwd(), filename), self.dir_root) logger_lpd_noaa.info("exit create_file") return
Open text file. Write one section at a time. Close text file. Move completed file to dir_root/noaa/ :return none:
entailment
def __write_top(self, filename_txt): """ Write the top section of the txt file. :param int section_num: Section number :return none: """ logger_lpd_noaa.info("writing section: {}".format("top")) self.__create_blanks("Top", self.noaa_data_sorted["Top"]) # Start writing the NOAA file section by section, starting at the very top of the template. self.noaa_txt += "# {}".format(self.noaa_data_sorted["Top"]['Study_Name']) self.__write_template_top() # We don't know what the full online resource path will be yet, so leave the base path only self.__write_k_v("Online_Resource", "{}/{}".format(self.wds_url, filename_txt), top=True) self.__write_k_v("Online_Resource_Description", " This file. NOAA WDS Paleo formatted metadata and data for version {} of this dataset.".format(self.version), indent=True) self.__write_k_v("Online_Resource", "{}".format(self.lpd_file_url), top=True) self.__write_k_v("Online_Resource_Description", " Linked Paleo Data (LiPD) formatted file containing the same metadata and data as this file, for version {} of this dataset.".format(self.version), indent=True) self.__write_k_v("Original_Source_URL", self.noaa_data_sorted["Top"]['Original_Source_URL'], top=True) self.noaa_txt += "\n# Description/Documentation lines begin with #\n# Data lines have no #\n#" self.__write_k_v("Archive", self.noaa_data_sorted["Top"]['Archive']) self.__write_k_v("Parameter_Keywords", self.noaa_data_sorted["Top"]['Parameter_Keywords']) # get the doi from the pub section of self.steps_dict[6] # doi = self.__get_doi() self.__write_k_v("Dataset_DOI", ', '.join(self.doi), False, True, False, False) # self.__write_k_v("Parameter_Keywords", self.steps_dict[section_num]['parameterKeywords']) self.__write_divider() logger_lpd_noaa.info("exit write_top") return
Write the top section of the txt file. :param int section_num: Section number :return none:
entailment
def __write_generic(self, header, d=None): """ Write a generic section to the .txt. This function is reused for multiple sections. :param str header: Header title for this section :param int section_num: :param dict d: Section from steps_dict :return none: """ logger_lpd_noaa.info("writing section: {}".format(header)) if not d: d = self.noaa_data_sorted[header] d = self.__create_blanks(header, d) self.__write_header_name(header) for key in NOAA_KEYS_BY_SECTION[header]: key = self.__get_noaa_key_w_context(header, key) # NOAA writes value and units on one line. Build the string here. # if key == 'coreLength': # value, unit = self.__get_corelength(val) # val = str(value) + " " + str(unit) # DOI id is nested in "identifier" block. Retrieve it. if key == "DOI": val = self.__get_doi(d) # Don't write out an empty DOI list. Write out empty string instead. if not val: val = "" elif key == "Investigators": # Investigators is a section by itself. "d" should be a direct list val = get_authors_as_str(d) # If we don't have investigator data, use the authors from the first publication entry if not val: val = self.__create_author_investigator_str() # lipd uses a singular "author" key, while NOAA uses plural "authors" key. elif key in ["Author", "Authors"]: # "d" has all publication data, so only pass through the d["Authors"] piece val = get_authors_as_str(d[key]) elif key == "Collection_Name": if not d[key]: # If there is not a collection name, then use the dsn so _something_ is there. val = self.dsn else: val = d[key] # Write the output line self.__write_k_v(str(self.__get_noaa_key(key)), val, indent=True) # Don't write a divider if there isn't a Chron section after species. It'll make a double. # if header == "Species" and not self.noaa_data_sorted["Species"]: # return self.__write_divider() return
Write a generic section to the .txt. This function is reused for multiple sections. :param str header: Header title for this section :param int section_num: :param dict d: Section from steps_dict :return none:
entailment
def __write_pub(self): """ Write pub section. There may be multiple, so write a generic section for each one. :return none: """ try: self.__reorganize_author() # Check all publications, and remove possible duplicate Full_Citations self.__rm_duplicate_citation_1() if not self.noaa_data_sorted["Publication"]: self.noaa_data_sorted["Publication"].append({"pubYear": ""}) for idx, pub in enumerate(self.noaa_data_sorted["Publication"]): logger_lpd_noaa.info("publication: {}".format(idx)) # Do not write out Data Citation publications. Check, and skip if necessary is_data_citation = self.__get_pub_type(pub) if not is_data_citation or is_data_citation and len(self.noaa_data_sorted["Publication"]) == 1: pub = self.__convert_keys_1("Publication", pub) self.__write_generic('Publication', pub) except KeyError: logger_lpd_noaa.info("write_pub: KeyError: pub section not found") except TypeError: logger_lpd_noaa.debug("write_pub: TypeError: pub not a list type") return
Write pub section. There may be multiple, so write a generic section for each one. :return none:
entailment
def __write_funding(self): """ Write funding section. There are likely multiple entries. :param dict d: :return none: """ self.__reorganize_funding() # if funding is empty, insert a blank entry so that it'll still write the empty section on the template. if not self.noaa_data_sorted["Funding_Agency"]: self.noaa_data_sorted["Funding_Agency"].append({"grant": "", "agency": ""}) for idx, entry in enumerate(self.noaa_data_sorted["Funding_Agency"]): logger_lpd_noaa.info("funding: {}".format(idx)) self.__write_generic('Funding_Agency', entry) return
Write funding section. There are likely multiple entries. :param dict d: :return none:
entailment
def __write_data(self, idx): """ Write out the measurement tables found in paleoData and chronData :return: """ pair = self.noaa_data_sorted["Data"][idx] # Run once for each pair (paleo+chron) of tables that was gathered earlier. # for idx, pair in enumerate(self.noaa_data_sorted["Data"]): lst_pc = ["chron", "paleo"] # loop once for paleo, once for chron for pc in lst_pc: # safeguard in case the table is an empty set. table = pair[pc] if pc == "paleo": self.__write_variables_1(table) self.__write_divider() self.__write_columns(pc, table) elif pc == "chron": # self.__write_variables_1(table) self.__write_columns(pc, table) self.__write_divider(nl=False) return
Write out the measurement tables found in paleoData and chronData :return:
entailment
def __write_variables_1(self, table): """ Retrieve variables from data table(s) and write to Variables section of txt file. :param dict table: Paleodata :return none: """ logger_lpd_noaa.info("writing section: {}".format("Variables")) # Write the template lines first self.__write_template_variable() try: self.noaa_txt += '#' # Special NOAA Request: Write the "year" column first always, if available # write year data first, when available for name, data in table["columns"].items(): if name == "year": # write first line in variables section here self.__write_variables_2(data) # leave the loop, because this is all we needed to accomplish break # all other cases for name, data in table["columns"].items(): # we already wrote out the year column, so don't duplicate. if name != "year": self.__write_variables_2(data) except KeyError as e: logger_lpd_noaa.warn("write_variables: KeyError: {} not found".format(e)) return
Retrieve variables from data table(s) and write to Variables section of txt file. :param dict table: Paleodata :return none:
entailment
def __write_variables_2(self, col): """ Use one column of data, to write one line of data in the variables section. :return none: """ col = self.__convert_keys_1("Variables", col) # Write one line for each column. One line has all metadata for one column. for entry in NOAA_KEYS_BY_SECTION["Variables"]: # May need a better way of handling this in the future. Need a strict list for this section. try: # First entry: Add extra hash and tab if entry == 'shortname': # DEPRECATED: Fixed spacing for variable names. # self.noaa_txt.write('{:<20}'.format('#' + str(col[entry]))) # Fluid spacing for variable names. Spacing dependent on length of variable names. self.noaa_txt += '{}\t'.format('#' + str(col[entry])) # Last entry: No space or comma elif entry == "additional": e = " " for item in ["notes", "uncertainty"]: try: if col[item]: e += str(col[item]).replace(",", ";") + "; " except KeyError: pass self.noaa_txt += '{} '.format(e) # elif entry == 'notes': # self.noaa_txt.write('{} '.format(str(col[entry]))) else: # This is for any entry that is not first or last in the line ordering # Account for nested entries. # if entry == "uncertainty": # try: # e = str(col["calibration"][entry]) # except KeyError: # e = "" if entry == "seasonality": try: e = str(col["climateInterpretation"][entry]) except KeyError: e = "" elif entry == "archive": e = self.noaa_data_sorted["Top"]["Archive"] elif entry == "dataType": # Lipd uses real data types (floats, ints), NOAA wants C or N (character or numeric) if col[entry] == "float": e = "N" else: e = "C" else: e = str(col[entry]) try: e = e.replace(",", ";") except AttributeError as ee: logger_lpd_noaa.warn("write_variables_2: AttributeError: {}, {}".format(e, ee)) self.noaa_txt += '{}, '.format(e) except KeyError as e: self.noaa_txt += '{:<0}'.format(',') logger_lpd_noaa.info("write_variables: KeyError: missing {}".format(e)) self.noaa_txt += '\n#' return
Use one column of data, to write one line of data in the variables section. :return none:
entailment
def __write_columns(self, pc, table): """ Read numeric data from csv and write to the bottom section of the txt file. :param dict table: Paleodata dictionary :return none: """ logger_lpd_noaa.info("writing section: data, csv values from file") # get filename for this table's csv data # filename = self.__get_filename(table) # logger_lpd_noaa.info("processing csv file: {}".format(filename)) # # get missing value for this table # # mv = self.__get_mv(table) # # write template lines # # self.__write_template_paleo(mv) if pc == "paleo": self.__write_template_paleo() elif pc == "chron": self.__write_template_chron() # continue if csv exists if self._values_exist(table): # logger_lpd_noaa.info("_write_columns: csv data exists: {}".format(filename)) # sort the dictionary so the year column is first _csv_data_by_name = self.__put_year_col_first(table["columns"]) # now split the sorted dictionary back into two lists (easier format to write to file) _names, _data = self.__rm_names_on_csv_cols(_csv_data_by_name) # write column variableNames self.__write_data_col_header(_names, pc) # write data columns index by index self.__write_data_col_vals(_data, pc) return
Read numeric data from csv and write to the bottom section of the txt file. :param dict table: Paleodata dictionary :return none:
entailment
def __write_k_v(self, k, v, top=False, bot=False, multi=False, indent=False): """ Write a key value pair to the output file. If v is a list, write multiple lines. :param k: Key :param v: Value :param bool top: Write preceding empty line :param bool bot: Write following empty line :param bool multi: v is a list :return none: """ if top: self.noaa_txt += "\n#" if multi: for item in v: if indent: self.noaa_txt += "\n# {}: {}".format(str(k), str(item)) else: self.noaa_txt += "\n# {}: {}".format(str(k), str(item)) else: if indent: self.noaa_txt += "\n# {}: {}".format(str(k), str(v)) else: self.noaa_txt += "\n# {}: {}".format(str(k), str(v)) if bot: self.noaa_txt += "\n#" return
Write a key value pair to the output file. If v is a list, write multiple lines. :param k: Key :param v: Value :param bool top: Write preceding empty line :param bool bot: Write following empty line :param bool multi: v is a list :return none:
entailment
def __write_divider(self, top=False, bot=False, nl=True): """ Write a divider line :return none: """ if top: self.noaa_txt += "\n#" if nl: self.noaa_txt += "\n" self.noaa_txt += "#------------------\n" if bot: self.noaa_txt += "\n#" return
Write a divider line :return none:
entailment
def __write_data_col_header(self, l, pc): """ Write the variableNames that are the column header in the "Data" section :param list l: variableNames :return none: """ count = len(l) if pc == "chron": self.noaa_txt += "# " for name in l: # last column - spacing not important if count == 1: self.noaa_txt += "{}\t".format(name) # all [:-1] columns - fixed spacing to preserve alignment else: self.noaa_txt += "{:<15}".format(name) count -= 1 self.noaa_txt += '\n'
Write the variableNames that are the column header in the "Data" section :param list l: variableNames :return none:
entailment
def __write_data_col_vals(self, ll, pc): """ Loop over value arrays and write index by index, to correspond to the rows of a txt file :param list ll: List of lists, column data :return: """ # all columns should have the same amount of values. grab that number try: _items_in_cols = len(ll[0]["values"]) for idx in range(0, _items_in_cols): # amount of columns _count = len(ll) self.noaa_txt += "# " for col in ll: self.noaa_txt += "{}\t".format(str(col["values"][idx])) _count -= 1 if (idx < _items_in_cols): self.noaa_txt += '\n' except IndexError: logger_lpd_noaa("_write_data_col_vals: IndexError: couldn't get length of columns") return
Loop over value arrays and write index by index, to correspond to the rows of a txt file :param list ll: List of lists, column data :return:
entailment
def copy(app, need, needs, option, need_id=None): """ Copies the value of one need option to another .. code-block:: jinja .. req:: copy-example :id: copy_1 :tags: tag_1, tag_2, tag_3 :status: open .. spec:: copy-example implementation :id: copy_2 :status: [[copy("status", "copy_1")]] :links: copy_1 :comment: [[copy("id")]] Copies status of ``copy_1`` to own status. Sets also a comment, which copies the id of own need. .. test:: test of specification and requirement :id: copy_3 :links: copy_2; [[copy('links', 'copy_2')]] :tags: [[copy('tags', 'copy_1')]] Set own link to ``copy_2`` and also copies all links from it. Also copies all tags from copy_1. .. req:: copy-example :id: copy_1 :tags: tag_1, tag_2, tag_3 :status: open .. spec:: copy-example implementation :id: copy_2 :status: [[copy("status", "copy_1")]] :links: copy_1 :comment: [[copy("id")]] Copies status of ``copy_1`` to own status. Sets also a comment, which copies the id of own need. .. test:: test of specification and requirement :id: copy_3 :links: copy_2; [[copy('links', 'copy_2')]] :tags: [[copy('tags', 'copy_1')]] Set own link to ``copy_2`` and also copies all links from it. Also copies all tags from copy_1. :param option: Name of the option to copy :param need_id: id of the need, which contains the source option. If None, current need is taken :return: string of copied need option """ if need_id is not None: need = needs[need_id] return need[option]
Copies the value of one need option to another .. code-block:: jinja .. req:: copy-example :id: copy_1 :tags: tag_1, tag_2, tag_3 :status: open .. spec:: copy-example implementation :id: copy_2 :status: [[copy("status", "copy_1")]] :links: copy_1 :comment: [[copy("id")]] Copies status of ``copy_1`` to own status. Sets also a comment, which copies the id of own need. .. test:: test of specification and requirement :id: copy_3 :links: copy_2; [[copy('links', 'copy_2')]] :tags: [[copy('tags', 'copy_1')]] Set own link to ``copy_2`` and also copies all links from it. Also copies all tags from copy_1. .. req:: copy-example :id: copy_1 :tags: tag_1, tag_2, tag_3 :status: open .. spec:: copy-example implementation :id: copy_2 :status: [[copy("status", "copy_1")]] :links: copy_1 :comment: [[copy("id")]] Copies status of ``copy_1`` to own status. Sets also a comment, which copies the id of own need. .. test:: test of specification and requirement :id: copy_3 :links: copy_2; [[copy('links', 'copy_2')]] :tags: [[copy('tags', 'copy_1')]] Set own link to ``copy_2`` and also copies all links from it. Also copies all tags from copy_1. :param option: Name of the option to copy :param need_id: id of the need, which contains the source option. If None, current need is taken :return: string of copied need option
entailment
def check_linked_values(app, need, needs, result, search_option, search_value, filter_string=None, one_hit=False): """ Returns a specific value, if for all linked needs a given option has a given value. The linked needs can be filtered by using the ``filter`` option. If ``one_hit`` is set to True, only one linked need must have a positive match for the searched value. **Examples** **Needs used as input data** .. code-block:: jinja .. req:: Input A :id: clv_A :status: in progress .. req:: Input B :id: clv_B :status: in progress .. spec:: Input C :id: clv_C :status: closed .. req:: Input A :id: clv_A :status: in progress :collapse: False .. req:: Input B :id: clv_B :status: in progress :collapse: False .. spec:: Input C :id: clv_C :status: closed :collapse: False **Example 1: Positive check** Status gets set to *progress*. .. code-block:: jinja .. spec:: result 1: Positive check :links: clv_A, clv_B :status: [[check_linked_values('progress', 'status', 'in progress' )]] .. spec:: result 1: Positive check :id: clv_1 :links: clv_A, clv_B :status: [[check_linked_values('progress', 'status', 'in progress' )]] :collapse: False **Example 2: Negative check** Status gets not set to *progress*, because status of linked need *clv_C* does not match *"in progress"*. .. code-block:: jinja .. spec:: result 2: Negative check :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress' )]] .. spec:: result 2: Negative check :id: clv_2 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress' )]] :collapse: False **Example 3: Positive check thanks of used filter** status gets set to *progress*, because linked need *clv_C* is not part of the filter. .. code-block:: jinja .. spec:: result 3: Positive check thanks of used filter :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', 'type == "req" ' )]] .. spec:: result 3: Positive check thanks of used filter :id: clv_3 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', 'type == "req" ' )]] :collapse: False **Example 4: Positive check thanks of one_hit option** Even *clv_C* has not the searched status, status gets anyway set to *progress*. That's because ``one_hit`` is used so that only one linked need must have the searched value. .. code-block:: jinja .. spec:: result 4: Positive check thanks of one_hit option :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] .. spec:: result 4: Positive check thanks of one_hit option :id: clv_4 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] :collapse: False **Result 5: Two checks and a joint status** Two checks are performed and both are positive. So their results get joined. .. code-block:: jinja .. spec:: result 5: Two checks and a joint status :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] [[check_linked_values('closed', 'status', 'closed', one_hit=True )]] .. spec:: result 5: Two checks and a joint status :id: clv_5 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] [[check_linked_values('closed', 'status', 'closed', one_hit=True )]] :collapse: False :param result: value, which gets returned if all linked needs have parsed the checks :param search_option: option name, which is used n linked needs for the search :param search_value: value, which an option of a linked need must match :param filter_string: Checks are only performed on linked needs, which pass the defined filter :param one_hit: If True, only one linked need must have a positive check :return: result, if all checks are positive """ links = need["links"] if not isinstance(search_value, list): search_value = [search_value] for link in links: if filter_string is not None: try: if not filter_single_need(needs[link], filter_string): continue except Exception as e: logger.warning("CheckLinkedValues: Filter {0} not valid: Error: {1}".format(filter_string, e)) if not one_hit and not needs[link][search_option] in search_value: return None elif one_hit and needs[link][search_option] in search_value: return result return result
Returns a specific value, if for all linked needs a given option has a given value. The linked needs can be filtered by using the ``filter`` option. If ``one_hit`` is set to True, only one linked need must have a positive match for the searched value. **Examples** **Needs used as input data** .. code-block:: jinja .. req:: Input A :id: clv_A :status: in progress .. req:: Input B :id: clv_B :status: in progress .. spec:: Input C :id: clv_C :status: closed .. req:: Input A :id: clv_A :status: in progress :collapse: False .. req:: Input B :id: clv_B :status: in progress :collapse: False .. spec:: Input C :id: clv_C :status: closed :collapse: False **Example 1: Positive check** Status gets set to *progress*. .. code-block:: jinja .. spec:: result 1: Positive check :links: clv_A, clv_B :status: [[check_linked_values('progress', 'status', 'in progress' )]] .. spec:: result 1: Positive check :id: clv_1 :links: clv_A, clv_B :status: [[check_linked_values('progress', 'status', 'in progress' )]] :collapse: False **Example 2: Negative check** Status gets not set to *progress*, because status of linked need *clv_C* does not match *"in progress"*. .. code-block:: jinja .. spec:: result 2: Negative check :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress' )]] .. spec:: result 2: Negative check :id: clv_2 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress' )]] :collapse: False **Example 3: Positive check thanks of used filter** status gets set to *progress*, because linked need *clv_C* is not part of the filter. .. code-block:: jinja .. spec:: result 3: Positive check thanks of used filter :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', 'type == "req" ' )]] .. spec:: result 3: Positive check thanks of used filter :id: clv_3 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', 'type == "req" ' )]] :collapse: False **Example 4: Positive check thanks of one_hit option** Even *clv_C* has not the searched status, status gets anyway set to *progress*. That's because ``one_hit`` is used so that only one linked need must have the searched value. .. code-block:: jinja .. spec:: result 4: Positive check thanks of one_hit option :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] .. spec:: result 4: Positive check thanks of one_hit option :id: clv_4 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] :collapse: False **Result 5: Two checks and a joint status** Two checks are performed and both are positive. So their results get joined. .. code-block:: jinja .. spec:: result 5: Two checks and a joint status :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] [[check_linked_values('closed', 'status', 'closed', one_hit=True )]] .. spec:: result 5: Two checks and a joint status :id: clv_5 :links: clv_A, clv_B, clv_C :status: [[check_linked_values('progress', 'status', 'in progress', one_hit=True )]] [[check_linked_values('closed', 'status', 'closed', one_hit=True )]] :collapse: False :param result: value, which gets returned if all linked needs have parsed the checks :param search_option: option name, which is used n linked needs for the search :param search_value: value, which an option of a linked need must match :param filter_string: Checks are only performed on linked needs, which pass the defined filter :param one_hit: If True, only one linked need must have a positive check :return: result, if all checks are positive
entailment
def calc_sum(app, need, needs, option, filter=None, links_only=False): """ Sums the values of a given option in filtered needs up to single number. Useful e.g. for calculating the amount of needed hours for implementation of all linked specification needs. **Input data** .. spec:: Do this :id: sum_input_1 :hours: 7 :collapse: False .. spec:: Do that :id: sum_input_2 :hours: 15 :collapse: False .. spec:: Do too much :id: sum_input_3 :hours: 110 :collapse: False **Example 2** .. code-block:: jinja .. req:: Result 1 :amount: [[calc_sum("hours")]] .. req:: Result 1 :amount: [[calc_sum("hours")]] :collapse: False **Example 2** .. code-block:: jinja .. req:: Result 2 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10")]] .. req:: Result 2 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10")]] :collapse: False **Example 3** .. code-block:: jinja .. req:: Result 3 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", links_only="True")]] .. req:: Result 3 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", links_only="True")]] :collapse: False **Example 4** .. code-block:: jinja .. req:: Result 4 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10", "True")]] .. req:: Result 4 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10", "True")]] :collapse: False :param option: Options, from which the numbers shall be taken :param filter: Filter string, which all needs must passed to get their value added. :param links_only: If "True", only linked needs are taken into account. :return: A float number """ if not links_only: check_needs = needs.values() else: check_needs = [] for link in need["links"]: check_needs.append(needs[link]) calculated_sum = 0 for check_need in check_needs: if filter is not None: try: if not filter_single_need(check_need, filter): continue except ValueError as e: pass except NeedInvalidFilter as ex: logger.warning('Given filter is not valid. Error: {}'.format(ex)) try: calculated_sum += float(check_need[option]) except ValueError: pass return calculated_sum
Sums the values of a given option in filtered needs up to single number. Useful e.g. for calculating the amount of needed hours for implementation of all linked specification needs. **Input data** .. spec:: Do this :id: sum_input_1 :hours: 7 :collapse: False .. spec:: Do that :id: sum_input_2 :hours: 15 :collapse: False .. spec:: Do too much :id: sum_input_3 :hours: 110 :collapse: False **Example 2** .. code-block:: jinja .. req:: Result 1 :amount: [[calc_sum("hours")]] .. req:: Result 1 :amount: [[calc_sum("hours")]] :collapse: False **Example 2** .. code-block:: jinja .. req:: Result 2 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10")]] .. req:: Result 2 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10")]] :collapse: False **Example 3** .. code-block:: jinja .. req:: Result 3 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", links_only="True")]] .. req:: Result 3 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", links_only="True")]] :collapse: False **Example 4** .. code-block:: jinja .. req:: Result 4 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10", "True")]] .. req:: Result 4 :links: sum_input_1; sum_input_3 :amount: [[calc_sum("hours", "hours.isdigit() and float(hours) > 10", "True")]] :collapse: False :param option: Options, from which the numbers shall be taken :param filter: Filter string, which all needs must passed to get their value added. :param links_only: If "True", only linked needs are taken into account. :return: A float number
entailment
def noaa_prompt(): """ Convert between NOAA and LiPD file formats. :return: """ logger_noaa.info("enter noaa") # Run lpd_noaa or noaa_lpd ? print("Which conversion?\n1. LPD to NOAA\n2. NOAA to LPD\n") mode = input("Option: ") logger_noaa.info("chose option: {}".format(mode)) return mode
Convert between NOAA and LiPD file formats. :return:
entailment
def noaa_prompt_1(): """ For converting LiPD files to NOAA, we need a couple more pieces of information to create the WDS links :return str _project: Project name :return float _version: Version number """ print("Enter the project information below. We'll use this to create the WDS URL") print("What is the project name?") _project = input(">") print("What is the project version?") _version = input(">") return _project, _version
For converting LiPD files to NOAA, we need a couple more pieces of information to create the WDS links :return str _project: Project name :return float _version: Version number
entailment
def noaa_to_lpd(files): """ Convert NOAA format to LiPD format :param dict files: Files metadata :return None: """ logger_noaa.info("enter process_noaa") # only continue if the user selected a mode correctly logger_noaa.info("Found {} NOAA txt file(s)".format(str(len(files[".txt"])))) print("Found {} NOAA txt file(s)".format(str(len(files[".txt"])))) # Process each available file of the specified .lpd or .txt type for file in files[".txt"]: # try to filter out example files and stuff without real data if "template" not in file["filename_ext"] and "example" not in file["filename_ext"]: os.chdir(file["dir"]) print('processing: {}'.format(file["filename_ext"])) logger_noaa.info("processing: {}".format(file["filename_ext"])) # Unzip file and get tmp directory path dir_tmp = create_tmp_dir() try: NOAA_LPD(file["dir"], dir_tmp, file["filename_no_ext"]).main() except Exception as e: print("Error: Unable to convert file: {}, {}".format(file["filename_no_ext"], e)) # Create the lipd archive in the original file's directory. zipper(root_dir=dir_tmp, name="bag", path_name_ext=os.path.join(file["dir"], file["filename_no_ext"] + ".lpd")) # Delete tmp folder and all contents os.chdir(file["dir"]) try: shutil.rmtree(dir_tmp) except FileNotFoundError: # directory is already gone. keep going. pass logger_noaa.info("exit noaa_to_lpd") return
Convert NOAA format to LiPD format :param dict files: Files metadata :return None:
entailment
def lpd_to_noaa(D, wds_url, lpd_url, version, path=""): """ Convert a LiPD format to NOAA format :param dict D: Metadata :return dict D: Metadata """ logger_noaa.info("enter process_lpd") d = D try: dsn = get_dsn(D) # Remove all the characters that are not allowed here. Since we're making URLs, they have to be compliant. dsn = re.sub(r'[^A-Za-z-.0-9]', '', dsn) # project = re.sub(r'[^A-Za-z-.0-9]', '', project) version = re.sub(r'[^A-Za-z-.0-9]', '', version) # Create the conversion object, and start the conversion process _convert_obj = LPD_NOAA(D, dsn, wds_url, lpd_url, version, path) _convert_obj.main() # get our new, modified master JSON from the conversion object d = _convert_obj.get_master() noaas = _convert_obj.get_noaa_texts() __write_noaas(noaas, path) # remove any root level urls that are deprecated d = __rm_wdc_url(d) except Exception as e: logger_noaa.error("lpd_to_noaa: {}".format(e)) print("Error: lpd_to_noaa: {}".format(e)) # logger_noaa.info("exit lpd_to_noaa") return d
Convert a LiPD format to NOAA format :param dict D: Metadata :return dict D: Metadata
entailment
def __write_noaas(dat, path): """ Use the filename - text data pairs to write the data as NOAA text files :param dict dat: NOAA data to be written :return none: """ for filename, text in dat.items(): try: with open(os.path.join(path, filename), "w+") as f: f.write(text) except Exception as e: print("write_noaas: There was a problem writing the NOAA text file: {}: {}".format(filename, e)) return
Use the filename - text data pairs to write the data as NOAA text files :param dict dat: NOAA data to be written :return none:
entailment
def _parse_java_version(line: str) -> tuple: """ Return the version number found in the first line of `java -version` >>> _parse_java_version('openjdk version "11.0.2" 2018-10-16') (11, 0, 2) """ m = VERSION_RE.search(line) version_str = m and m.group(0).replace('"', '') or '0.0.0' if '_' in version_str: fst, snd = version_str.split('_', maxsplit=2) version = parse_version(fst) return (version[1], version[2], int(snd)) else: return parse_version(version_str)
Return the version number found in the first line of `java -version` >>> _parse_java_version('openjdk version "11.0.2" 2018-10-16') (11, 0, 2)
entailment
def find_java_home(cratedb_version: tuple) -> str: """ Return a path to a JAVA_HOME suites for the given CrateDB version """ if MIN_VERSION_FOR_JVM11 <= cratedb_version < (4, 0): # Supports 8 to 11+, use whatever is set return os.environ.get('JAVA_HOME', '') if cratedb_version < MIN_VERSION_FOR_JVM11: return _find_matching_java_home(lambda ver: ver[0] == 8) else: return _find_matching_java_home(lambda ver: ver[0] >= 11)
Return a path to a JAVA_HOME suites for the given CrateDB version
entailment
def zipper(root_dir="", name="", path_name_ext=""): """ Zips up directory back to the original location :param str root_dir: Root directory of the archive :param str name: <datasetname>.lpd :param str path_name_ext: /path/to/filename.lpd """ logger_zips.info("re_zip: name: {}, dir_tmp: {}".format(path_name_ext, root_dir)) # creates a zip archive in current directory. "somefile.lpd.zip" shutil.make_archive(path_name_ext, format='zip', root_dir=root_dir, base_dir=name) # drop the .zip extension. only keep .lpd os.rename("{}.zip".format(path_name_ext), path_name_ext) return
Zips up directory back to the original location :param str root_dir: Root directory of the archive :param str name: <datasetname>.lpd :param str path_name_ext: /path/to/filename.lpd
entailment
def unzipper(filename, dir_tmp): """ Unzip .lpd file contents to tmp directory. :param str filename: filename.lpd :param str dir_tmp: Tmp folder to extract contents to :return None: """ logger_zips.info("enter unzip") # Unzip contents to the tmp directory try: with zipfile.ZipFile(filename) as f: f.extractall(dir_tmp) except FileNotFoundError as e: logger_zips.debug("unzip: FileNotFound: {}, {}".format(filename, e)) shutil.rmtree(dir_tmp) logger_zips.info("exit unzip") return
Unzip .lpd file contents to tmp directory. :param str filename: filename.lpd :param str dir_tmp: Tmp folder to extract contents to :return None:
entailment
def percentile(sorted_values, p): """Calculate the percentile using the nearest rank method. >>> percentile([15, 20, 35, 40, 50], 50) 35 >>> percentile([15, 20, 35, 40, 50], 40) 20 >>> percentile([], 90) Traceback (most recent call last): ... ValueError: Too few data points (0) for 90th percentile """ size = len(sorted_values) idx = (p / 100.0) * size - 0.5 if idx < 0 or idx > size: raise ValueError('Too few data points ({}) for {}th percentile'.format(size, p)) return sorted_values[int(idx)]
Calculate the percentile using the nearest rank method. >>> percentile([15, 20, 35, 40, 50], 50) 35 >>> percentile([15, 20, 35, 40, 50], 40) 20 >>> percentile([], 90) Traceback (most recent call last): ... ValueError: Too few data points (0) for 90th percentile
entailment
def get_sampler(sample_mode: str): """Return a sampler constructor >>> get_sampler('all') <class 'cr8.metrics.All'> >>> get_sampler('reservoir') <class 'cr8.metrics.UniformReservoir'> >>> get_sampler('reservoir:100') functools.partial(<class 'cr8.metrics.UniformReservoir'>, size=100) """ if sample_mode == 'all': return All mode = sample_mode.split(':') if mode[0] == 'reservoir': if len(mode) == 2: return partial(UniformReservoir, size=int(mode[1])) else: return UniformReservoir raise TypeError(f'Invalid sample_mode: {sample_mode}')
Return a sampler constructor >>> get_sampler('all') <class 'cr8.metrics.All'> >>> get_sampler('reservoir') <class 'cr8.metrics.UniformReservoir'> >>> get_sampler('reservoir:100') functools.partial(<class 'cr8.metrics.UniformReservoir'>, size=100)
entailment
def query(self, criteria, **opts): """ Returns a dict of the query, including the results :param critera: string of search criteria :param **opts: :formats: json/xml (default: json) :timezone: timezone to use (default: UTC) :time_from: 15m ago from now (datetime) :time_to: right now (datetime) """ time_now = datetime.datetime.now().replace(second=0, microsecond=0) right_now = time_now.isoformat() minutes_ago = (time_now - datetime.timedelta(minutes=15)).isoformat() formats = opts.get('formats', 'json') timezone = opts.get('timezone', 'UTC') time_from = opts.get('time_from', minutes_ago) time_to = opts.get('time_to', right_now) # setting up options t_options = { 'q': criteria, 'format': formats, 'tz': timezone, 'from': time_from, 'to': time_to, } options = '&'.join(['{}={}'.format(k, v) for k, v in t_options.items()]) req = requests.get('%s?%s' % (self.url, options), auth=self.auth) try: data = req.json() except json.decoder.JSONDecodeError: data = [] return { 'data': data, 'response': req.status_code, 'reason': req.reason, }
Returns a dict of the query, including the results :param critera: string of search criteria :param **opts: :formats: json/xml (default: json) :timezone: timezone to use (default: UTC) :time_from: 15m ago from now (datetime) :time_to: right now (datetime)
entailment
def excel_main(file): """ Parse data from Excel spreadsheets into LiPD files. :param dict file: File metadata (source, name, etc) :return str dsn: Dataset name """ os.chdir(file["dir"]) name_ext = file["filename_ext"] # Filename without extension name = file["filename_no_ext"] # remove foreign characters to prevent wiki uploading errors name = normalize_name(name) print("processing: {}".format(name_ext)) logger_excel.info("processing: {}".format(name_ext)) pending_csv = [] final = OrderedDict() logger_excel.info("variables initialized") # Create a temporary folder and set paths dir_tmp = create_tmp_dir() """ EACH DATA TABLE WILL BE STRUCTURED LIKE THIS "paleo_chron": "paleo", "pc_idx": 1, "model_idx": "", "table_type": "measurement", "table_idx": 1, "name": sheet, "filename": sheet, "data": { column data } """ # Open excel workbook with filename try: workbook = xlrd.open_workbook(name_ext) logger_excel.info("opened XLRD workbook") except Exception as e: # There was a problem opening a file with XLRD print("Failed to open Excel workbook: {}".format(name)) workbook = None logger_excel.debug("excel: xlrd failed to open workbook: {}, {}".format(name, e)) if workbook: # Build sheets, but don't make full filenames yet. Need to parse metadata sheet first to get datasetname. sheets, ct_paleo, ct_chron, metadata_str = _get_sheet_metadata(workbook, name) # METADATA WORKSHEETS # Parse Metadata sheet and add to output dictionary if metadata_str: logger_excel.info("parsing worksheet: {}".format(metadata_str)) final = cells_dn_meta(workbook, metadata_str, 0, 0, final) # Now that we have the dataSetName, we can use it to build filenames dsn = __get_datasetname(final, name) filename = str(dsn) + ".lpd" sheets = __set_sheet_filenames(sheets, dsn) dir_bag = os.path.join(dir_tmp, "bag") dir_data = os.path.join(dir_bag, 'data') # Make folders in tmp os.mkdir(os.path.join(dir_bag)) os.mkdir(os.path.join(dir_data)) # PALEO AND CHRON SHEETS for sheet in sheets: logger_excel.info("parsing data worksheet: {}".format(sheet["new_name"])) sheet_meta, sheet_csv = _parse_sheet(workbook, sheet) if sheet_csv and sheet_meta: pending_csv.append(sheet_csv) sheet["data"] = sheet_meta # create the metadata skeleton where we will place the tables. dynamically add empty table blocks for data. skeleton_paleo, skeleton_chron = _create_skeleton_1(sheets) # Reorganize sheet metadata into LiPD structure d_paleo, d_chron = _place_tables_main(sheets, skeleton_paleo, skeleton_chron) # Add organized metadata into final dictionary final['paleoData'] = d_paleo final['chronData'] = d_chron # OUTPUT # Create new files and dump data in dir_data os.chdir(dir_data) # WRITE CSV _write_data_csv(pending_csv) # JSON-LD # Invoke DOI Resolver Class to update publisher data try: logger_excel.info("invoking doi resolver") final = DOIResolver(file["dir"], name, final).main() except Exception as e: print("Error: doi resolver failed: {}".format(name)) logger_excel.debug("excel: doi resolver failed: {}, {}".format(name, e)) # Dump final_dict to a json file. final["lipdVersion"] = 1.2 final["createdBy"] = "excel" write_json_to_file(final) # Move files to bag root for re-bagging # dir : dir_data -> dir_bag logger_excel.info("start cleanup") dir_cleanup(dir_bag, dir_data) # Create a bag for the 3 files finish_bag(dir_bag) # dir: dir_tmp -> dir_root os.chdir(file["dir"]) # Check if same lpd file exists. If so, delete so new one can be made if os.path.isfile(filename): os.remove(filename) # Zip dir_bag. Creates in dir_root directory logger_excel.info("re-zip and rename") zipper(root_dir=dir_tmp, name="bag", path_name_ext=os.path.join(file["dir"], filename)) # Move back to dir_root for next loop. os.chdir(file["dir"]) # Cleanup and remove tmp directory shutil.rmtree(dir_tmp) return dsn
Parse data from Excel spreadsheets into LiPD files. :param dict file: File metadata (source, name, etc) :return str dsn: Dataset name
entailment
def _get_sheet_metadata(workbook, name): """ Get worksheet metadata. The sheet names tell us what type of table it is and where in the LiPD structure the data should be placed. Example VALID sheet name: paleo1measurement1 paleo1model1ensemble1 paleo1model1distribution1 Example INVALID sheet names: paleo1measurement paleo1ensemble paleo1_measurement1 NOTE: since each model will only have one ensemble and one summary, they should not have a trailing index number. :param obj workbook: Excel workbook :param str name: Dataset name :return dict int int str: """ ct_paleo = 1 ct_chron = 1 metadata_str = "" sheets = [] skip_sheets = ["example", "sample", "lists", "guidelines"] # Check what worksheets are available, so we know how to proceed. for sheet in workbook.sheet_names(): # Use this for when we are dealing with older naming styles of "data (qc), data, and chronology" old = "".join(sheet.lower().strip().split()) # Don't parse example sheets. If these words are in the sheet, assume we skip them. if not any(word in sheet.lower() for word in skip_sheets): # Group the related sheets together, so it's easier to place in the metadata later. if 'metadata' in sheet.lower(): metadata_str = sheet # Skip the 'about' and 'proxy' sheets altogether. Proceed with all other sheets. elif "about" not in sheet.lower() and "proxy" not in sheet.lower(): logger_excel.info("creating sheets metadata") # If this is a valid sheet name, we will receive a regex object back. m = re.match(re_sheet, sheet.lower()) # Valid regex object. This is a valid sheet name and we can use that to build the sheet metadata. if m: sheets, paleo_ct, chron_ct = _sheet_meta_from_regex(m, sheets, sheet, name, ct_paleo, ct_chron) # Older excel template style: backwards compatibility. Hard coded for one sheet per table. elif old == "data" or "data(qc)" in old or "data(original)" in old: sheets.append({ "paleo_chron": "paleo", "idx_pc": ct_paleo, "idx_model": None, "table_type": "measurement", "idx_table": 1, "old_name": sheet, "new_name": sheet, "filename": "paleo{}measurementTable1.csv".format(ct_paleo), "table_name": "paleo{}measurementTable1".format(ct_paleo), "data": "" }) ct_paleo += 1 # Older excel template style: backwards compatibility. Hard coded for one sheet per table. elif old == "chronology": sheets.append({ "paleo_chron": "chron", "idx_pc": ct_chron, "idx_model": None, "table_type": "measurement", "idx_table": 1, "old_name": sheet, "new_name": sheet, "filename": "chron{}measurementTable1.csv".format(ct_chron), "table_name": "chron{}measurementTable1".format(ct_chron), "data": "" }) ct_chron += 1 else: # Sheet name does not conform to standard. Guide user to create a standardized sheet name. print("This sheet name does not conform to naming standard: {}".format(sheet)) sheets, paleo_ct, chron_ct = _sheet_meta_from_prompts(sheets, sheet, name, ct_paleo, ct_chron) return sheets, ct_paleo, ct_chron, metadata_str
Get worksheet metadata. The sheet names tell us what type of table it is and where in the LiPD structure the data should be placed. Example VALID sheet name: paleo1measurement1 paleo1model1ensemble1 paleo1model1distribution1 Example INVALID sheet names: paleo1measurement paleo1ensemble paleo1_measurement1 NOTE: since each model will only have one ensemble and one summary, they should not have a trailing index number. :param obj workbook: Excel workbook :param str name: Dataset name :return dict int int str:
entailment
def _sheet_meta_from_prompts(sheets, old_name, name, ct_paleo, ct_chron): """ Guide the user to create a proper, standardized sheet name :param list sheets: Running list of sheet metadata :param str old_name: Original sheet name :param str name: Data set name :param int ct_paleo: Running count of paleoData tables :param int ct_chron: Running count of chronData tables :return sheets paleo_ct chron_ct: Updated sheets and counts """ cont = True # Loop until valid sheet name is built, or user gives up while cont: try: pc = input("Is this a (p)aleo or (c)hronology sheet?").lower() if pc in ("p", "c", "paleo", "chron", "chronology"): tt = input("Is this a (d)istribution, (e)nsemble, (m)easurement, or (s)ummary sheet?").lower() if tt in EXCEL_SHEET_TYPES["distribution"] or tt in EXCEL_SHEET_TYPES["ensemble"] \ or tt in EXCEL_SHEET_TYPES["summary"] or tt in EXCEL_SHEET_TYPES["measurement"]: # valid answer, keep going if tt in EXCEL_SHEET_TYPES["distribution"]: tt = "distribution" elif tt in EXCEL_SHEET_TYPES["summary"]: tt = "summary" elif tt in EXCEL_SHEET_TYPES["ensemble"]: tt = "ensemble" elif tt in EXCEL_SHEET_TYPES["measurement"]: tt = "measurement" if pc in EXCEL_SHEET_TYPES["paleo"]: if tt in ["ensemble", "summary"]: sheet = "{}{}{}{}".format("paleo", ct_paleo, tt, 1) else: sheet = "{}{}{}".format("paleo", ct_paleo, tt) elif pc in EXCEL_SHEET_TYPES["chron"]: if tt in ["ensemble", "summary"]: sheet = "{}{}{}{}".format("chron", ct_chron, tt, 1) else: sheet = "{}{}{}".format("chron", ct_chron, tt) # Test the sheet that was built from the user responses. # If it matches the Regex, then continue to build the sheet metadata. If not, try again or skip sheet. m = re.match(re_sheet, sheet.lower()) if m: sheets, ct_paleo, ct_chron = _sheet_meta_from_regex(m, sheets, old_name, name, ct_paleo, ct_chron) print("Sheet created: {}".format(sheet)) cont = False else: resp = input("invalid sheet name. try again? (y/n): ") if resp == "n": print("No valid sheet name was created. Skipping sheet: {}".format(sheet)) cont = False except Exception as e: logger_excel.debug("excel: sheet_meta_from_prompts: error during prompts, {}".format(e)) cont = False print("=====================================================") return sheets, ct_paleo, ct_chron
Guide the user to create a proper, standardized sheet name :param list sheets: Running list of sheet metadata :param str old_name: Original sheet name :param str name: Data set name :param int ct_paleo: Running count of paleoData tables :param int ct_chron: Running count of chronData tables :return sheets paleo_ct chron_ct: Updated sheets and counts
entailment
def _sheet_meta_from_regex(m, sheets, old_name, name, ct_paleo, ct_chron): """ Build metadata for a sheet. Receive valid regex match object and use that to create metadata. :param obj m: Regex match object :param list sheets: Running list of sheet metadata :param str old_name: Original sheet name :param str name: Data set name :param int ct_paleo: Running count of paleoData tables :param int ct_chron: Running count of chronData tables :return sheets paleo_ct chron_ct: Updated sheets and counts """ try: idx_model = None idx_table = None pc = m.group(1) # Get the model idx number from string if it exists if m.group(3): idx_model = int(m.group(4)) # check if there's an index (for distribution tables) if m.group(6): idx_table = int(m.group(6)) # find out table type if pc == "paleodata" or pc == "paleo": pc = "paleo" ct_paleo += 1 elif pc == "chrondata" or pc == "chron": pc = "chron" ct_chron += 1 # build filename and table name strings. build table name first, then make filename from the table_name new_name = "{}{}".format(pc, m.group(2)) if idx_model: new_name = "{}model{}".format(new_name, idx_model) new_name = "{}{}".format(new_name, m.group(5)) if idx_table: new_name = "{}{}".format(new_name, m.group(6)) filename = "{}.csv".format(new_name) except Exception as e: logger_excel.debug("excel: sheet_meta_from_regex: error during setup, {}".format(e)) # Standard naming. This matches the regex and the sheet name is how we want it. # paleo/chron - idx - table_type - idx # Not sure what to do with m.group(2) yet # ex: m.groups() = [ paleo, 1, model, 1, ensemble, 1 ] try: sheets.append({ "old_name": old_name, "new_name": new_name, "filename": filename, "paleo_chron": pc, "idx_pc": int(m.group(2)), "idx_model": idx_model, "idx_table": idx_table, "table_type": m.group(5), "data": "" }) except Exception as e: print("error: build sheets") logger_excel.debug("excel: build_sheet: unable to build sheet, {}".format(e)) return sheets, ct_paleo, ct_chron
Build metadata for a sheet. Receive valid regex match object and use that to create metadata. :param obj m: Regex match object :param list sheets: Running list of sheet metadata :param str old_name: Original sheet name :param str name: Data set name :param int ct_paleo: Running count of paleoData tables :param int ct_chron: Running count of chronData tables :return sheets paleo_ct chron_ct: Updated sheets and counts
entailment
def _place_tables_section(skeleton_section, sheet, keys_section): """ Place data into skeleton for either a paleo or chron section. :param dict skeleton_section: Empty or current progress of skeleton w/ data :param dict sheet: Sheet metadata :param list keys_section: Paleo or Chron specific keys :return dict: Skeleton section full of data """ logger_excel.info("enter place_tables_section") try: logger_excel.info("excel: place_tables_section: placing table: {}".format(sheet["new_name"])) new_name = sheet["new_name"] logger_excel.info("placing_tables_section: {}".format(new_name)) # get all the sheet metadata needed for this function idx_pc = sheet["idx_pc"] - 1 idx_model = sheet["idx_model"] idx_table = sheet["idx_table"] table_type = sheet["table_type"] data = sheet["data"] # paleoMeas or chronMeas key key_1 = keys_section[0] # paleoModel or chronModel key key_2 = keys_section[1] # Is this a measurement, or distribution table? if idx_table: # Yes, a table idx exists, so decrement it. idx_table = sheet["idx_table"] - 1 # Is this a ensemble, dist, or summary table? if idx_model: # Yes, a model idx exists, so decrement it. idx_model -= 1 except Exception as e: logger_excel.debug("excel: place_tables_section: error during setup, {}".format(e)) # If it's measurement table, it goes in first. try: if table_type == "measurement": skeleton_section[idx_pc][key_1][idx_table] = data # Other types of tables go one step below elif table_type in ["ensemble", "distribution", "summary"]: if table_type == "summary": skeleton_section[idx_pc][key_2][idx_model]["summaryTable"] = data elif table_type == "ensemble": skeleton_section[idx_pc][key_2][idx_model]["ensembleTable"] = data elif table_type == "distribution": skeleton_section[idx_pc][key_2][idx_model]["distributionTable"][idx_table] = data except Exception as e: logger_excel.warn("excel: place_tables_section: Unable to place table {}, {}".format(new_name, e)) logger_excel.info("exit place_tables_section") return skeleton_section
Place data into skeleton for either a paleo or chron section. :param dict skeleton_section: Empty or current progress of skeleton w/ data :param dict sheet: Sheet metadata :param list keys_section: Paleo or Chron specific keys :return dict: Skeleton section full of data
entailment
def _place_tables_main(sheets, skeleton_paleo, skeleton_chron): """ All the data has been parsed, skeletons have been created, now put the data into the skeletons. :param list sheets: All metadata needed to place sheet data into the LiPD structure :param list skeleton_paleo: The empty skeleton where we will place data :param list skeleton_chron: The empty skeleton where we will place data :return: """ logger_excel.info("enter place_tables_main") for sheet in sheets: pc = sheet["paleo_chron"] if pc == "paleo": skeleton_paleo = _place_tables_section(skeleton_paleo, sheet, ["paleoMeasurementTable", "paleoModel"]) elif pc == "chron": skeleton_chron = _place_tables_section(skeleton_chron, sheet, ["chronMeasurementTable", "chronModel"]) # when returning, these should no longer be skeletons. They should be tables filled with data logger_excel.info("exit place_tables_main") return skeleton_paleo, skeleton_chron
All the data has been parsed, skeletons have been created, now put the data into the skeletons. :param list sheets: All metadata needed to place sheet data into the LiPD structure :param list skeleton_paleo: The empty skeleton where we will place data :param list skeleton_chron: The empty skeleton where we will place data :return:
entailment
def _get_table_counts(sheet, num_section): """ Loop through sheet metadata and count how many of each table type is needed at each index. Example: 'paleo 1' needs {'2 measurement tables', '1 model table, with 1 summary, 1 ensemble, and 3 distributions'} :param dict sheet: Sheet metadata :param dict num_section: Rolling number counts of table types for each index. :return: """ tt = sheet["table_type"] idx_pc = sheet["idx_pc"] idx_table = sheet["idx_table"] idx_model = sheet["idx_model"] # Have we started counters for this idx model yet?? if idx_pc not in num_section: # No, create the counters and start tracking table counts num_section[idx_pc] = {"ct_meas": 0, "ct_model": 0, "ct_in_model": {}} # Compare indices to get the highest index number for this table type # If we have a higher number model index, then increment out models count try: # Is this a ens, dist, or summary table? if idx_model: # Yes it is. # Is this model idx higher than ours? if idx_model > num_section[idx_pc]["ct_model"]: # Yes. Now, we have N number of model tables. num_section[idx_pc]["ct_model"] = idx_model # Have we started counters for this idx model yet?? if idx_model not in num_section[idx_pc]["ct_in_model"]: # No, create the counters and start tracking table counts. num_section[idx_pc]["ct_in_model"][idx_model] = {"ct_ens": 0, "ct_sum": 0, "ct_dist": 0} except Exception as e: logger_excel.debug("excel: get_table_counts: error incrementing model counts, ".format(e)) # Incrementer! # For the given table type, track the highest index number. # That number is how many tables we need to make of that type. Ex. 'measurement4', we need to make 4 empty tables try: if tt == "measurement": # Is this meas table a higher idx? if idx_table > num_section[idx_pc]["ct_meas"]: # Yes, set this idx to the counter. num_section[idx_pc]["ct_meas"] = idx_table elif tt == "distribution": # Is this dist table a higher idx? if idx_table > num_section[idx_pc]["ct_in_model"][idx_model]["ct_dist"]: # Yes, set this idx to the counter. num_section[idx_pc]["ct_in_model"][idx_model]["ct_dist"] = idx_table elif tt == "summary": # Summary tables are not indexed. Only one per table. num_section[idx_pc]["ct_in_model"][idx_model]["ct_sum"] = 1 elif tt == "ensemble": # Ensemble tables are not indexed. Only one per table. num_section[idx_pc]["ct_in_model"][idx_model]["ct_ens"] = 1 except Exception as e: logger_excel.debug("excel: get_table_counts: error incrementing table count".format(e)) return num_section
Loop through sheet metadata and count how many of each table type is needed at each index. Example: 'paleo 1' needs {'2 measurement tables', '1 model table, with 1 summary, 1 ensemble, and 3 distributions'} :param dict sheet: Sheet metadata :param dict num_section: Rolling number counts of table types for each index. :return:
entailment
def _create_skeleton_3(pc, l, num_section): """ Bottom level: {"measurement": [], "model": [{summary, distributions, ensemble}]} Fill in measurement and model tables with N number of EMPTY meas, summary, ensemble, and distributions. :param str pc: Paleo or Chron "mode" :param list l: :param dict num_section: :return dict: """ logger_excel.info("enter create_skeleton_inner_2") # Table Template: Model template_model = {"summaryTable": {}, "ensembleTable": {}, "distributionTable": []} # Build string appropriate for paleo/chron mode pc_meas = "{}MeasurementTable".format(pc) pc_mod = "{}Model".format(pc) # Loop for each table count for idx1, table in num_section.items(): try: # Create N number of empty measurement lists l[idx1 - 1][pc_meas] = [None] * num_section[idx1]["ct_meas"] # Create N number of empty model table lists l[idx1 - 1][pc_mod] = [copy.deepcopy(template_model)] * num_section[idx1]["ct_model"] # Create N number of empty model tables at list index # for idx2, nums in table["ct_in_model"].items(): dists = [] try: # Create N number of empty distributions at list index [dists.append({}) for i in range(0, nums["ct_dist"])] except IndexError as e: logger_excel.debug("excel: create_metadata_skeleton: paleo tables messed up, {}".format(e)) # Model template complete, insert it at list index l[idx1 - 1][pc_mod][idx2-1] = {"summaryTable": {}, "ensembleTable": {}, "distributionTable": dists} except IndexError as e: logger_excel.warn("create_skeleton_inner_tables: IndexError: {}".format(e)) except KeyError as e: logger_excel.warn("create_skeleton_inner_tables: KeyError: {}".format(e)) return l
Bottom level: {"measurement": [], "model": [{summary, distributions, ensemble}]} Fill in measurement and model tables with N number of EMPTY meas, summary, ensemble, and distributions. :param str pc: Paleo or Chron "mode" :param list l: :param dict num_section: :return dict:
entailment
def _create_skeleton_2(l, pc, num_section, template): """ Mid level: {"measurement", "model"} Fill in paleoData tables with N number of EMPTY measurement and models. :param list l: paleoData or chronData list of tables :param dict num_section: Number of tables needed for each table type :param dict template: The empty template for the measurement and model top section. :return list: paleoData or chronData list of tables - full skeleton """ logger_excel.info("enter create_skeleton_inner_1") try: # Create N number of paleoData/chronData tables. l = [copy.deepcopy(template)] * len(num_section) # Create the necessary tables inside of "model" l = _create_skeleton_3(pc, l, num_section) except Exception as e: logger_excel.warn("excel: create_skeleton_inner_main: error duplicating template tables, {}".format(e)) return l
Mid level: {"measurement", "model"} Fill in paleoData tables with N number of EMPTY measurement and models. :param list l: paleoData or chronData list of tables :param dict num_section: Number of tables needed for each table type :param dict template: The empty template for the measurement and model top section. :return list: paleoData or chronData list of tables - full skeleton
entailment
def _create_skeleton_1(sheets): """ Top level: {"chronData", "paleoData"} Fill in paleoData/chronData tables with N number of EMPTY measurement and models. :return list: Blank list of N indices """ logger_excel.info("enter create_skeleton_main") # Table template: paleoData template_paleo = {"paleoMeasurementTable": [], "paleoModel": []} template_chron = {"chronMeasurementTable": [], "chronModel": []} num_chron = {} num_paleo = {} paleo = [] chron = [] # Get table counts for all table types. # Table types: paleoData, chronData, model, meas, dist, summary, ensemble for sheet in sheets: pc = sheet["paleo_chron"] # Tree for chron types if pc == "chron": num_chron = _get_table_counts(sheet, num_chron) elif pc == "paleo": num_paleo = _get_table_counts(sheet, num_paleo) # Create metadata skeleton for using out table counts. paleo = _create_skeleton_2(paleo, "paleo", num_paleo, template_paleo) chron = _create_skeleton_2(chron, "chron", num_chron, template_chron) logger_excel.info("exit create_skeleton_main") return paleo, chron
Top level: {"chronData", "paleoData"} Fill in paleoData/chronData tables with N number of EMPTY measurement and models. :return list: Blank list of N indices
entailment
def _parse_sheet(workbook, sheet): """ The universal spreadsheet parser. Parse chron or paleo tables of type ensemble/model/summary. :param str name: Filename :param obj workbook: Excel Workbook :param dict sheet: Sheet path and naming info :return dict dict: Table metadata and numeric data """ logger_excel.info("enter parse_sheet: {}".format(sheet["old_name"])) # Markers to track where we are on the sheet ensemble_on = False var_header_done = False metadata_on = False metadata_done = False data_on = False notes = False # Open the sheet from the workbook temp_sheet = workbook.sheet_by_name(sheet["old_name"]) filename = sheet["filename"] # Store table metadata and numeric data separately table_name = "{}DataTableName".format(sheet["paleo_chron"]) # Organize our root table data table_metadata = OrderedDict() table_metadata[table_name] = sheet["new_name"] table_metadata['filename'] = filename table_metadata['missingValue'] = 'nan' if "ensemble" in sheet["new_name"]: ensemble_on = True # Store all CSV in here by rows table_data = {filename: []} # Master list of all column metadata column_metadata = [] # Index tracks which cells are being parsed num_col = 0 num_row = 0 nrows = temp_sheet.nrows col_total = 0 # Tracks which "number" each metadata column is assigned col_add_ct = 1 header_keys = [] variable_keys = [] variable_keys_lower = [] mv = "" try: # Loop for every row in the sheet for i in range(0, nrows): # Hold the contents of the current cell cell = temp_sheet.cell_value(num_row, num_col) row = temp_sheet.row(num_row) # Skip all template lines if isinstance(cell, str): # Note and missing value entries are rogue. They are not close to the other data entries. if cell.lower().strip() not in EXCEL_TEMPLATE: if "notes" in cell.lower() and not metadata_on: # Store at the root table level nt = temp_sheet.cell_value(num_row, 1) if nt not in EXCEL_TEMPLATE: table_metadata["notes"] = nt elif cell.lower().strip() in ALTS_MV: # Store at the root table level and in our function mv = temp_sheet.cell_value(num_row, 1) # Add if not placeholder value if mv not in EXCEL_TEMPLATE: table_metadata["missingValue"] = mv # Variable template header row elif cell.lower() in EXCEL_HEADER and not metadata_on and not data_on: # Grab the header line row = temp_sheet.row(num_row) header_keys = _get_header_keys(row) # Turn on the marker var_header_done = True # Data section (bottom of sheet) elif data_on: # Parse the row, clean, and add to table_data table_data = _parse_sheet_data_row(temp_sheet, num_row, col_total, table_data, filename, mv) # Metadata section. (top) elif metadata_on: # Reached an empty cell while parsing metadata. Mark the end of the section. if cell in EMPTY: metadata_on = False metadata_done = True # Create a list of all the variable names found for entry in column_metadata: try: # var keys is used as the variableName entry in each column's metadata variable_keys.append(entry["variableName"].strip()) # var keys lower is used for comparing and finding the data header row variable_keys_lower.append(entry["variableName"].lower().strip()) except KeyError: # missing a variableName key pass # Not at the end of the section yet. Parse the metadata else: # Get the row data row = temp_sheet.row(num_row) # Get column metadata col_tmp = _compile_column_metadata(row, header_keys, col_add_ct) # Append to master list column_metadata.append(col_tmp) col_add_ct += 1 # Variable metadata, if variable header exists elif var_header_done and not metadata_done: # Start piecing column metadata together with their respective variable keys metadata_on = True # Get the row data row = temp_sheet.row(num_row) # Get column metadata col_tmp = _compile_column_metadata(row, header_keys, col_add_ct) # Append to master list column_metadata.append(col_tmp) col_add_ct += 1 # Variable metadata, if variable header does not exist elif not var_header_done and not metadata_done and cell: # LiPD Version 1.1 and earlier: Chronology sheets don't have variable headers # We could blindly parse, but without a header row_num we wouldn't know where # to save the metadata # Play it safe and assume data for first column only: variable name metadata_on = True # Get the row data row = temp_sheet.row(num_row) # Get column metadata col_tmp = _compile_column_metadata(row, header_keys, col_add_ct) # Append to master list column_metadata.append(col_tmp) col_add_ct += 1 # Data variable header row. Column metadata exists and metadata_done marker is on. # This is where we compare top section variableNames to bottom section variableNames to see if # we need to start parsing the column values else: try: # Clean up variable_keys_lower so we all variable names change from "age(yrs BP)" to "age" # Units in parenthesis make it too difficult to compare variables. Remove them. row = _rm_units_from_var_names_multi(row) if metadata_done and any(i in row for i in variable_keys_lower): data_on = True # Take the difference of the two lists. If anything exists, then that's a problem __compare_vars(row, variable_keys_lower, sheet["old_name"]) # Ensemble columns are counted differently. if ensemble_on: # Get the next row, and count the data cells. col_total = len(temp_sheet.row(num_row+1)) # If there's an empty row, between, then try the next row. if col_total < 2: col_total = temp_sheet.row(num_row + 2) try: ens_cols = [] [ens_cols.append(i+1) for i in range(0, col_total-1)] column_metadata[1]["number"] = ens_cols except IndexError: logger_excel.debug("excel: parse_sheet: unable to add ensemble 'number' key") except KeyError: logger_excel.debug("excel: parse_sheet: unable to add ensemble 'number' list at key") # All other cass, columns are the length of column_metadata else: col_total = len(column_metadata) except AttributeError: pass # cell is not a string, and lower() was not a valid call. # If this is a numeric cell, 99% chance it's parsing the data columns. elif isinstance(cell, float) or isinstance(cell, int): if data_on or metadata_done: # Parse the row, clean, and add to table_data table_data = _parse_sheet_data_row(temp_sheet, num_row, col_total, table_data, filename, mv) # Move on to the next row num_row += 1 table_metadata["columns"] = column_metadata except IndexError as e: logger_excel.debug("parse_sheet: IndexError: sheet: {}, row_num: {}, col_num: {}, {}".format(sheet, num_row, num_col, e)) # If there isn't any data in this sheet, and nothing was parsed, don't let this # move forward to final output. if not table_data[filename]: table_data = None table_metadata = None logger_excel.info("exit parse_sheet") return table_metadata, table_data
The universal spreadsheet parser. Parse chron or paleo tables of type ensemble/model/summary. :param str name: Filename :param obj workbook: Excel Workbook :param dict sheet: Sheet path and naming info :return dict dict: Table metadata and numeric data
entailment
def _parse_sheet_data_row(temp_sheet, num_row, col_total, table_data, filename, mv): """ Parse a row from the data section of the sheet. Add the cleaned row data to the overall table data. :param obj temp_sheet: Excel sheet :param int num_row: Current sheet row :param int col_total: Number of column variables in this sheet :param dict table_data: Running record of table data :param str filename: Filename for this table :param str mv: Missing value :return dict: Table data with appended row """ # Get row of data row = temp_sheet.row(num_row) # In case our row holds more cells than the amount of columns we have, slice the row # We don't want to have extra empty cells in our output. row = row[:col_total] # Replace missing values where necessary row = _replace_mvs(row, mv) # Append row to list we will use to write out csv file later. table_data[filename].append(row) return table_data
Parse a row from the data section of the sheet. Add the cleaned row data to the overall table data. :param obj temp_sheet: Excel sheet :param int num_row: Current sheet row :param int col_total: Number of column variables in this sheet :param dict table_data: Running record of table data :param str filename: Filename for this table :param str mv: Missing value :return dict: Table data with appended row
entailment
def _replace_mvs(row, mv): """ Replace Missing Values in the data rows where applicable :param list row: Row :return list: Modified row """ for idx, v in enumerate(row): try: if v.value.lower() in EMPTY or v.value.lower() == mv: row[idx] = "nan" else: row[idx] = v.value except AttributeError: if v.value == mv: row[idx] = "nan" else: row[idx] = v.value return row
Replace Missing Values in the data rows where applicable :param list row: Row :return list: Modified row
entailment
def _get_header_keys(row): """ Get the variable header keys from this special row :return list: Header keys """ # Swap out NOAA keys for LiPD keys for idx, key in enumerate(row): key_low = key.value.lower() # Simple case: Nothing fancy here, just map to the LiPD key counterpart. if key_low in EXCEL_LIPD_MAP_FLAT: row[idx] = EXCEL_LIPD_MAP_FLAT[key_low] # Nested data case: Check if this is a calibration, interpretation, or some other data that needs to be nested. # elif key_low: # pass # Unknown key case: Store the key as-is because we don't have a LiPD mapping for it. else: try: row[idx] = key.value except AttributeError as e: logger_excel.warn("excel_main: get_header_keys: unknown header key, unable to add: {}".format(e)) # Since we took a whole row of cells, we have to drop off the empty cells at the end of the row. header_keys = _rm_cells_reverse(row) return header_keys
Get the variable header keys from this special row :return list: Header keys
entailment
def _rm_units_from_var_name_single(var): """ NOTE: USE THIS FOR SINGLE CELLS ONLY When parsing sheets, all variable names be exact matches when cross-referenceing the metadata and data sections However, sometimes people like to put "age (years BP)" in one section, and "age" in the other. This causes problems. We're using this regex to match all variableName cells and remove the "(years BP)" where applicable. :param str var: Variable name :return str: Variable name """ # Use the regex to match the cell m = re.match(re_var_w_units, var) # Should always get a match, but be careful anyways. if m: # m.group(1): variableName # m.group(2): units in parenthesis (may not exist). try: var = m.group(1).strip().lower() # var = m.group(1).strip().lower() except Exception: # This must be a malformed cell somehow. This regex should match every variableName cell. # It didn't work out. Return the original var as a fallback pass return var
NOTE: USE THIS FOR SINGLE CELLS ONLY When parsing sheets, all variable names be exact matches when cross-referenceing the metadata and data sections However, sometimes people like to put "age (years BP)" in one section, and "age" in the other. This causes problems. We're using this regex to match all variableName cells and remove the "(years BP)" where applicable. :param str var: Variable name :return str: Variable name
entailment
def _rm_units_from_var_names_multi(row): """ Wrapper around "_rm_units_from_var_name_single" for doing a list instead of a single cell. :param list row: Variable names :return list: Variable names """ l2 = [] # Check each var in the row for idx, var in enumerate(row): l2.append(_rm_units_from_var_name_single(row[idx].value)) return l2
Wrapper around "_rm_units_from_var_name_single" for doing a list instead of a single cell. :param list row: Variable names :return list: Variable names
entailment
def _compile_interpretation(data): """ Compile the interpretation data into a list of multiples, based on the keys provided. Disassemble the key to figure out how to place the data :param dict data: Interpretation data (unsorted) :return dict: Interpretation data (sorted) """ # KEY FORMAT : "interpretation1_somekey" _count = 0 # Determine how many entries we are going to need, by checking the interpretation index in the string for _key in data.keys(): _key_low = _key.lower() # Get regex match m = re.match(re_interpretation, _key_low) # If regex match was successful.. if m: # Check if this interpretation count is higher than what we have. _curr_count = int(m.group(1)) if _curr_count > _count: # New max count, record it. _count = _curr_count # Create the empty list with X entries for the interpretation data _tmp = [{} for i in range(0, _count)] # Loop over all the interpretation keys and data for k, v in data.items(): # Get the resulting regex data. # EXAMPLE ENTRY: "interpretation1_variable" # REGEX RESULT: ["1", "variable"] m = re.match(re_interpretation, k) # Get the interpretation index number idx = int(m.group(1)) # Get the field variable key = m.group(2) # Place this data in the _tmp array. Remember to adjust given index number for 0-indexing _tmp[idx-1][key] = v # Return compiled interpretation data return _tmp
Compile the interpretation data into a list of multiples, based on the keys provided. Disassemble the key to figure out how to place the data :param dict data: Interpretation data (unsorted) :return dict: Interpretation data (sorted)
entailment
def _compile_column_metadata(row, keys, number): """ Compile column metadata from one excel row ("9 part data") :param list row: Row of cells :param list keys: Variable header keys :return dict: Column metadata """ # Store the variable keys by index in a dictionary _column = {} _interpretation = {} _calibration = {} _physical = {} # Use the header keys to place the column data in the dictionary if keys: for idx, key in enumerate(keys): _key_low = key.lower() # Special case: Calibration data if re.match(re_calibration, _key_low): m = re.match(re_calibration, _key_low) if m: _key = m.group(1) _calibration[_key] = row[idx].value # Special case: PhysicalSample data elif re.match(re_physical, _key_low): m = re.match(re_physical, _key_low) if m: _key = m.group(1) _physical[_key] = row[idx].value # Special case: Interpretation data elif re.match(re_interpretation, _key_low): # Put interpretation data in a tmp dictionary that we'll sort later. _interpretation[_key_low] = row[idx].value else: try: val = row[idx].value except Exception: logger_excel.info("compile_column_metadata: Couldn't get value from row cell") val = "n/a" try: if key == "variableName": val = _rm_units_from_var_name_single(row[idx].value) except Exception: # when a variableName fails to split, keep the name as-is and move on. pass _column[key] = val _column["number"] = number if _calibration: _column["calibration"] = _calibration # Only allow physicalSample on measured variableTypes. duh. if _physical and _column["variableType"] == "measured": _column["physicalSample"] = _physical if _interpretation: _interpretation_data = _compile_interpretation(_interpretation) _column["interpretation"] = _interpretation_data # If there are not keys, that means it's a header-less metadata section. else: # Assume we only have one cell, because we have no keys to know what data is here. try: val = row[0].value.lower() except AttributeError: val = row[0].value except Exception: logger_excel.info("compile_column_metadata: Couldn't get value from row cell") val = "n/a" val = _rm_units_from_var_name_single(val) _column["variableName"] = val _column["number"] = number # Add this column to the overall metadata, but skip if there's no data present _column = {k: v for k, v in _column.items() if v} return _column
Compile column metadata from one excel row ("9 part data") :param list row: Row of cells :param list keys: Variable header keys :return dict: Column metadata
entailment
def _rm_cells_reverse(l): """ Remove the cells that are empty or template in reverse order. Stop when you hit data. :param list l: One row from the spreadsheet :return list: Modified row """ rm = [] # Iter the list in reverse, and get rid of empty and template cells for idx, key in reversed(list(enumerate(l))): if key.lower() in EXCEL_TEMPLATE: rm.append(idx) elif key in EMPTY: rm.append(idx) else: break for idx in rm: l.pop(idx) return l
Remove the cells that are empty or template in reverse order. Stop when you hit data. :param list l: One row from the spreadsheet :return list: Modified row
entailment
def _write_data_csv(csv_data): """ CSV data has been parsed by this point, so take it and write it file by file. :return: """ logger_excel.info("enter write_data_csv") # Loop for each file and data that is stored for file in csv_data: for filename, data in file.items(): # Make sure we're working with the right data types before trying to open and write a file if isinstance(filename, str) and isinstance(data, list): try: with open(filename, 'w+') as f: w = csv.writer(f) for line in data: w.writerow(line) except Exception: logger_excel.debug("write_data_csv: Unable to open/write file: {}".format(filename)) logger_excel.info("exit write_data_csv") return
CSV data has been parsed by this point, so take it and write it file by file. :return:
entailment
def geometry_linestring(lat, lon, elev): """ GeoJSON Linestring. Latitude and Longitude have 2 values each. :param list lat: Latitude values :param list lon: Longitude values :return dict: """ logger_excel.info("enter geometry_linestring") d = OrderedDict() coordinates = [] temp = ["", ""] # Point type, Matching pairs. if lat[0] == lat[1] and lon[0] == lon[1]: logger_excel.info("matching geo coordinate") lat.pop() lon.pop() d = geometry_point(lat, lon, elev) else: # Creates coordinates list logger_excel.info("unique geo coordinates") for i in lon: temp[0] = i for j in lat: temp[1] = j coordinates.append(copy.copy(temp)) if elev: for i in coordinates: i.append(elev) # Create geometry block d['type'] = 'Linestring' d['coordinates'] = coordinates logger_excel.info("exit geometry_linestring") return d
GeoJSON Linestring. Latitude and Longitude have 2 values each. :param list lat: Latitude values :param list lon: Longitude values :return dict:
entailment
def geometry_range(crd_range, elev, crd_type): """ Range of coordinates. (e.g. 2 latitude coordinates, and 0 longitude coordinates) :param crd_range: Latitude or Longitude values :param elev: Elevation value :param crd_type: Coordinate type, lat or lon :return dict: """ d = OrderedDict() coordinates = [[] for i in range(len(crd_range))] # latitude if crd_type == "lat": for idx, i in enumerate(crd_range): coordinates[idx] = [crd_range[idx], "nan"] if elev: coordinates[idx].append(elev) # longitude elif crd_type == "lon": for idx, i in enumerate(crd_range): coordinates[idx] = ["nan", crd_range[idx]] if elev: coordinates[idx].append(elev) d["type"] = "Range" d["coordinates"] = coordinates return d
Range of coordinates. (e.g. 2 latitude coordinates, and 0 longitude coordinates) :param crd_range: Latitude or Longitude values :param elev: Elevation value :param crd_type: Coordinate type, lat or lon :return dict:
entailment
def geometry_point(lat, lon, elev): """ GeoJSON point. Latitude and Longitude only have one value each :param list lat: Latitude values :param list lon: Longitude values :param float elev: Elevation value :return dict: """ logger_excel.info("enter geometry_point") coordinates = [] point_dict = OrderedDict() for idx, val in enumerate(lat): try: coordinates.append(lon[idx]) coordinates.append(lat[idx]) except IndexError as e: print("Error: Invalid geo coordinates") logger_excel.debug("geometry_point: IndexError: lat: {}, lon: {}, {}".format(lat, lon, e)) coordinates.append(elev) point_dict['type'] = 'Point' point_dict['coordinates'] = coordinates logger_excel.info("exit geometry_point") return point_dict
GeoJSON point. Latitude and Longitude only have one value each :param list lat: Latitude values :param list lon: Longitude values :param float elev: Elevation value :return dict:
entailment
def compile_geometry(lat, lon, elev): """ Take in lists of lat and lon coordinates, and determine what geometry to create :param list lat: Latitude values :param list lon: Longitude values :param float elev: Elevation value :return dict: """ logger_excel.info("enter compile_geometry") lat = _remove_geo_placeholders(lat) lon = _remove_geo_placeholders(lon) # 4 coordinate values if len(lat) == 2 and len(lon) == 2: logger_excel.info("found 4 coordinates") geo_dict = geometry_linestring(lat, lon, elev) # # 4 coordinate values # if (lat[0] != lat[1]) and (lon[0] != lon[1]): # geo_dict = geometry_polygon(lat, lon) # # 3 unique coordinates # else: # geo_dict = geometry_multipoint(lat, lon) # # 2 coordinate values elif len(lat) == 1 and len(lon) == 1: logger_excel.info("found 2 coordinates") geo_dict = geometry_point(lat, lon, elev) # coordinate range. one value given but not the other. elif (None in lon and None not in lat) or (len(lat) > 0 and len(lon) == 0): geo_dict = geometry_range(lat, elev, "lat") elif (None in lat and None not in lon) or (len(lon) > 0 and len(lat) == 0): geo_dict = geometry_range(lat, elev, "lon") # Too many points, or no points else: geo_dict = {} logger_excel.warn("compile_geometry: invalid coordinates: lat: {}, lon: {}".format(lat, lon)) logger_excel.info("exit compile_geometry") return geo_dict
Take in lists of lat and lon coordinates, and determine what geometry to create :param list lat: Latitude values :param list lon: Longitude values :param float elev: Elevation value :return dict:
entailment
def compile_geo(d): """ Compile top-level Geography dictionary. :param d: :return: """ logger_excel.info("enter compile_geo") d2 = OrderedDict() # get max number of sites, or number of coordinate points given. num_loc = _get_num_locations(d) # if there's one more than one location put it in a collection if num_loc > 1: d2["type"] = "FeatureCollection" features = [] for idx in range(0, num_loc): # Do process for one site site = _parse_geo_locations(d, idx) features.append(site) d2["features"] = features # if there's only one location elif num_loc == 1: d2 = _parse_geo_location(d) logger_excel.info("exit compile_geo") return d2
Compile top-level Geography dictionary. :param d: :return:
entailment
def _get_num_locations(d): """ Find out how many locations are being parsed. Compare lengths of each coordinate list and return the max :param dict d: Geo metadata :return int: Max number of locations """ lengths = [] for key in EXCEL_GEO: try: if key != "siteName": lengths.append(len(d[key])) except Exception: lengths.append(1) try: num = max(lengths) except ValueError: num = 0 return num
Find out how many locations are being parsed. Compare lengths of each coordinate list and return the max :param dict d: Geo metadata :return int: Max number of locations
entailment
def _parse_geo_location(d): """ Parse one geo location :param d: :return: """ d2 = OrderedDict() filt = {} d2['type'] = 'Feature' # If the necessary keys are missing, put in placeholders so there's no KeyErrors. for key in EXCEL_GEO: if key not in d: d[key] = "" # Compile the geometry based on the info available. d2['geometry'] = compile_geometry([d['latMin'], d['latMax']], [d['lonMin'], d['lonMax']], d['elevation']) d2['properties'] = {'siteName': d['siteName']} return d2
Parse one geo location :param d: :return:
entailment
def _parse_geo_locations(d, idx): """ Parse one geo location :param d: :return: """ d2 = OrderedDict() filt = {} d2['type'] = 'Feature' # If the necessary keys are missing, put in placeholders so there's no KeyErrors. for key in EXCEL_GEO: if key not in d: d[key] = "" for key in EXCEL_GEO: try: if key == "siteName" and isinstance(d["siteName"], str): filt["siteName"] = d["siteName"] else: filt[key] = d[key][idx] except KeyError: filt[key] = None except TypeError: filt[key] = None # Compile the geometry based on the info available. d2['geometry'] = compile_geometry([filt['latMin'], filt['latMax']], [filt['lonMin'], filt['lonMax']], filt['elevation']) d2['properties'] = {'siteName': filt['siteName']} return d2
Parse one geo location :param d: :return:
entailment
def compile_authors(cell): """ Split the string of author names into the BibJSON format. :param str cell: Data from author cell :return: (list of dicts) Author names """ logger_excel.info("enter compile_authors") author_lst = [] s = cell.split(';') for w in s: author_lst.append(w.lstrip()) logger_excel.info("exit compile_authors") return author_lst
Split the string of author names into the BibJSON format. :param str cell: Data from author cell :return: (list of dicts) Author names
entailment
def compile_temp(d, key, value): """ Compiles temporary dictionaries for metadata. Adds a new entry to an existing dictionary. :param dict d: :param str key: :param any value: :return dict: """ if not value: d[key] = None elif len(value) == 1: d[key] = value[0] else: d[key] = value return d
Compiles temporary dictionaries for metadata. Adds a new entry to an existing dictionary. :param dict d: :param str key: :param any value: :return dict:
entailment
def compile_fund(workbook, sheet, row, col): """ Compile funding entries. Iter both rows at the same time. Keep adding entries until both cells are empty. :param obj workbook: :param str sheet: :param int row: :param int col: :return list of dict: l """ logger_excel.info("enter compile_fund") l = [] temp_sheet = workbook.sheet_by_name(sheet) while col < temp_sheet.ncols: col += 1 try: # Make a dictionary for this funding entry. _curr = { 'agency': temp_sheet.cell_value(row, col), 'grant': temp_sheet.cell_value(row+1, col), "principalInvestigator": temp_sheet.cell_value(row+2, col), "country": temp_sheet.cell_value(row + 3, col) } # Make a list for all _exist = [temp_sheet.cell_value(row, col), temp_sheet.cell_value(row+1, col), temp_sheet.cell_value(row+2, col), temp_sheet.cell_value(row+3, col)] # Remove all empty items from the list _exist = [i for i in _exist if i] # If we have all empty entries, then don't continue. Quit funding and return what we have. if not _exist: return l # We have funding data. Add this funding block to the growing list. l.append(_curr) except IndexError as e: logger_excel.debug("compile_fund: IndexError: sheet:{} row:{} col:{}, {}".format(sheet, row, col, e)) logger_excel.info("exit compile_fund") return l
Compile funding entries. Iter both rows at the same time. Keep adding entries until both cells are empty. :param obj workbook: :param str sheet: :param int row: :param int col: :return list of dict: l
entailment
def __compare_vars(a1, a2, name): """ Check that the metadata variable names are the same as the data header variable names :param list a1: Variable names :param list a2: Variable names :param str name: Sheet name :return bool: Truth """ try: a1 = [i for i in a1 if i] a2 = [i for i in a2 if i] a3 = set(a1).symmetric_difference(set(a2)) if a3: print("- Error: Variables are not entered correctly in sheet: {}\n\tUnmatched variables: {}".format(name, a3)) except Exception as e: logger_excel.error("compare_vars: {}".format(e)) return
Check that the metadata variable names are the same as the data header variable names :param list a1: Variable names :param list a2: Variable names :param str name: Sheet name :return bool: Truth
entailment
def __get_datasetname(d, filename): """ Get the filename based on the dataset name in the metadata :param str filename: Filename.lpd :return str: Filename """ try: filename = d["dataSetName"] except KeyError: logger_excel.info("get_datasetname: KeyError: No dataSetName found. Reverting to: {}".format(filename)) return filename
Get the filename based on the dataset name in the metadata :param str filename: Filename.lpd :return str: Filename
entailment
def __set_sheet_filenames(sheets, n): """ Use the dataset name to build the filenames in the sheets metadata :param list sheets: Sheet metadata :param str n: Dataset Name :return list: Sheet metadata """ try: for idx, sheet in enumerate(sheets): try: sheets[idx]["filename"] = "{}.{}".format(n, sheet["filename"]) except Exception as e: logger_excel.error("set_sheet_filenames: inner: {}".format(e), exc_info=True) except Exception as q: logger_excel.error("set_sheet_filenames: outer: {}".format(q), exc_info=True) return sheets
Use the dataset name to build the filenames in the sheets metadata :param list sheets: Sheet metadata :param str n: Dataset Name :return list: Sheet metadata
entailment
def name_to_jsonld(title_in): """ Convert formal titles to camelcase json_ld text that matches our context file Keep a growing list of all titles that are being used in the json_ld context :param str title_in: :return str: """ title_out = '' try: title_in = title_in.lower() title_out = EXCEL_LIPD_MAP_FLAT[title_in] except (KeyError, AttributeError) as e: if "(" in title_in: title_in = title_in.split("(")[0].strip() # try to find an exact match first. try: v = EXCEL_LIPD_MAP_FLAT[title_in] return v except KeyError: pass # if no exact match, find whatever is a closest match for k, v in EXCEL_LIPD_MAP_FLAT.items(): if k in title_in: return v if not title_out: logger_excel.debug("name_to_jsonld: No match found: {}".format(title_in)) return title_out
Convert formal titles to camelcase json_ld text that matches our context file Keep a growing list of all titles that are being used in the json_ld context :param str title_in: :return str:
entailment
def instance_str(cell): """ Match data type and return string :param any cell: :return str: """ if isinstance(cell, str): return 'str' elif isinstance(cell, int): return 'int' elif isinstance(cell, float): return 'float' else: return 'unknown'
Match data type and return string :param any cell: :return str:
entailment
def extract_units(string_in): """ Extract units from parenthesis in a string. i.e. "elevation (meters)" :param str string_in: :return str: """ start = '(' stop = ')' return string_in[string_in.index(start) + 1:string_in.index(stop)]
Extract units from parenthesis in a string. i.e. "elevation (meters)" :param str string_in: :return str:
entailment
def cells_rt_meta_pub(workbook, sheet, row, col, pub_qty): """ Publication section is special. It's possible there's more than one publication. :param obj workbook: :param str sheet: :param int row: :param int col: :param int pub_qty: Number of distinct publication sections in this file :return list: Cell data for a specific row """ logger_excel.info("enter cells_rt_meta_pub") col_loop = 0 cell_data = [] temp_sheet = workbook.sheet_by_name(sheet) while col_loop < pub_qty: col += 1 col_loop += 1 cell_data.append(temp_sheet.cell_value(row, col)) logger_excel.info("exit cells_rt_meta_pub") return cell_data
Publication section is special. It's possible there's more than one publication. :param obj workbook: :param str sheet: :param int row: :param int col: :param int pub_qty: Number of distinct publication sections in this file :return list: Cell data for a specific row
entailment
def cells_rt_meta(workbook, sheet, row, col): """ Traverse all cells in a row. If you find new data in a cell, add it to the list. :param obj workbook: :param str sheet: :param int row: :param int col: :return list: Cell data for a specific row """ logger_excel.info("enter cells_rt_meta") col_loop = 0 cell_data = [] temp_sheet = workbook.sheet_by_name(sheet) while col_loop < temp_sheet.ncols: col += 1 col_loop += 1 try: if temp_sheet.cell_value(row, col) != xlrd.empty_cell and temp_sheet.cell_value(row, col) != '': cell_data.append(temp_sheet.cell_value(row, col)) except IndexError as e: logger_excel.warn("cells_rt_meta: IndexError: sheet: {}, row: {}, col: {}, {}".format(sheet, row, col, e)) logger_excel.info("exit cells_right_meta") return cell_data
Traverse all cells in a row. If you find new data in a cell, add it to the list. :param obj workbook: :param str sheet: :param int row: :param int col: :return list: Cell data for a specific row
entailment
def cells_dn_meta(workbook, sheet, row, col, final_dict): """ Traverse all cells in a column moving downward. Primarily created for the metadata sheet, but may use elsewhere. Check the cell title, and switch it to. :param obj workbook: :param str sheet: :param int row: :param int col: :param dict final_dict: :return: none """ logger_excel.info("enter cells_dn_meta") row_loop = 0 pub_cases = ['id', 'year', 'author', 'journal', 'issue', 'volume', 'title', 'pages', 'reportNumber', 'abstract', 'alternateCitation'] geo_cases = ['latMin', 'lonMin', 'lonMax', 'latMax', 'elevation', 'siteName', 'location'] funding_cases = ["agency", "grant", "principalInvestigator", "country"] # Temp pub_qty = 0 geo_temp = {} general_temp = {} pub_temp = [] funding_temp = [] temp_sheet = workbook.sheet_by_name(sheet) # Loop until we hit the max rows in the sheet while row_loop < temp_sheet.nrows: try: # Get cell value cell = temp_sheet.cell_value(row, col) # If there is content in the cell... if cell not in EMPTY: # Convert title to correct format, and grab the cell data for that row title_formal = temp_sheet.cell_value(row, col) title_json = name_to_jsonld(title_formal) # If we don't have a title for it, then it's not information we want to grab if title_json: # Geo if title_json in geo_cases: cell_data = cells_rt_meta(workbook, sheet, row, col) geo_temp = compile_temp(geo_temp, title_json, cell_data) # Pub # Create a list of dicts. One for each pub column. elif title_json in pub_cases: # Authors seem to be the only consistent field we can rely on to determine number of Pubs. if title_json == 'author': cell_data = cells_rt_meta(workbook, sheet, row, col) pub_qty = len(cell_data) for i in range(pub_qty): author_lst = compile_authors(cell_data[i]) pub_temp.append({'author': author_lst, 'pubDataUrl': 'Manually Entered'}) else: cell_data = cells_rt_meta_pub(workbook, sheet, row, col, pub_qty) for pub in range(pub_qty): if title_json == 'id': pub_temp[pub]['identifier'] = [{"type": "doi", "id": cell_data[pub]}] else: pub_temp[pub][title_json] = cell_data[pub] # Funding elif title_json in funding_cases: if title_json == "agency": funding_temp = compile_fund(workbook, sheet, row, col) # All other cases do not need fancy structuring else: cell_data = cells_rt_meta(workbook, sheet, row, col) general_temp = compile_temp(general_temp, title_json, cell_data) except IndexError as e: logger_excel.debug("cells_dn_datasheets: IndexError: sheet: {}, row: {}, col: {}, {}".format(sheet, row, col, e)) row += 1 row_loop += 1 # Compile the more complicated items geo = compile_geo(geo_temp) logger_excel.info("compile metadata dictionary") # Insert into final dictionary final_dict['@context'] = "context.jsonld" final_dict['pub'] = pub_temp final_dict['funding'] = funding_temp final_dict['geo'] = geo # Add remaining general items for k, v in general_temp.items(): final_dict[k] = v logger_excel.info("exit cells_dn_meta") return final_dict
Traverse all cells in a column moving downward. Primarily created for the metadata sheet, but may use elsewhere. Check the cell title, and switch it to. :param obj workbook: :param str sheet: :param int row: :param int col: :param dict final_dict: :return: none
entailment
def count_chron_variables(temp_sheet): """ Count the number of chron variables :param obj temp_sheet: :return int: variable count """ total_count = 0 start_row = traverse_to_chron_var(temp_sheet) while temp_sheet.cell_value(start_row, 0) != '': total_count += 1 start_row += 1 return total_count
Count the number of chron variables :param obj temp_sheet: :return int: variable count
entailment
def get_chron_var(temp_sheet, start_row): """ Capture all the vars in the chron sheet (for json-ld output) :param obj temp_sheet: :param int start_row: :return: (list of dict) column data """ col_dict = OrderedDict() out_list = [] column = 1 while (temp_sheet.cell_value(start_row, 0) != '') and (start_row < temp_sheet.nrows): short_cell = temp_sheet.cell_value(start_row, 0) units_cell = temp_sheet.cell_value(start_row, 1) long_cell = temp_sheet.cell_value(start_row, 2) # Fill the dictionary for this column col_dict['number'] = column col_dict['variableName'] = short_cell col_dict['description'] = long_cell col_dict['units'] = units_cell out_list.append(col_dict.copy()) start_row += 1 column += 1 return out_list
Capture all the vars in the chron sheet (for json-ld output) :param obj temp_sheet: :param int start_row: :return: (list of dict) column data
entailment
def traverse_to_chron_data(temp_sheet): """ Traverse down to the first row that has chron data :param obj temp_sheet: :return int: traverse_row """ traverse_row = traverse_to_chron_var(temp_sheet) reference_var = temp_sheet.cell_value(traverse_row, 0) # Traverse past all the short_names, until you hit a blank cell (the barrier) while temp_sheet.cell_value(traverse_row, 0) != '': traverse_row += 1 # Traverse past the empty cells until we hit the chron data area while temp_sheet.cell_value(traverse_row, 0) == '': traverse_row += 1 # Check if there is a header row. If there is, move past it. We don't want that data if temp_sheet.cell_value(traverse_row, 0) == reference_var: traverse_row += 1 logger_excel.info("traverse_to_chron_data: row:{}".format(traverse_row)) return traverse_row
Traverse down to the first row that has chron data :param obj temp_sheet: :return int: traverse_row
entailment
def traverse_to_chron_var(temp_sheet): """ Traverse down to the row that has the first variable :param obj temp_sheet: :return int: """ row = 0 while row < temp_sheet.nrows - 1: if 'Parameter' in temp_sheet.cell_value(row, 0): row += 1 break row += 1 logger_excel.info("traverse_to_chron_var: row:{}".format(row)) return row
Traverse down to the row that has the first variable :param obj temp_sheet: :return int:
entailment
def get_chron_data(temp_sheet, row, total_vars): """ Capture all data in for a specific chron data row (for csv output) :param obj temp_sheet: :param int row: :param int total_vars: :return list: data_row """ data_row = [] missing_val_list = ['none', 'na', '', '-'] for i in range(0, total_vars): cell = temp_sheet.cell_value(row, i) if isinstance(cell, str): cell = cell.lower() if cell in missing_val_list: cell = 'nan' data_row.append(cell) return data_row
Capture all data in for a specific chron data row (for csv output) :param obj temp_sheet: :param int row: :param int total_vars: :return list: data_row
entailment
def _remove_geo_placeholders(l): """ Remove placeholders from coordinate lists and sort :param list l: Lat or long list :return list: Modified list """ vals = [] for i in l: if isinstance(i, list): for k in i: if isinstance(k, float) or isinstance(k, int): vals.append(k) elif isinstance(i, float) or isinstance(i, int): vals.append(i) vals.sort() return vals
Remove placeholders from coordinate lists and sort :param list l: Lat or long list :return list: Modified list
entailment
def read_jsonld(): """ Find jsonld file in the cwd (or within a 2 levels below cwd), and load it in. :return dict: Jsonld data """ _d = {} try: # Find a jsonld file in cwd. If none, fallback for a json file. If neither found, return empty. _filename = [file for file in os.listdir() if file.endswith(".jsonld")][0] if not _filename: _filename = [file for file in os.listdir() if file.endswith(".json")][0] if _filename: try: # Load and decode _d = demjson.decode_file(_filename, decode_float=float) logger_jsons.info("Read JSONLD successful: {}".format(_filename)) except FileNotFoundError as fnf: print("Error: metadata file not found: {}".format(_filename)) logger_jsons.error("read_jsonld: FileNotFound: {}, {}".format(_filename, fnf)) except Exception: try: _d = demjson.decode_file(_filename, decode_float=float, encoding="latin-1") logger_jsons.info("Read JSONLD successful: {}".format(_filename)) except Exception as e: print("Error: unable to read metadata file: {}".format(e)) logger_jsons.error("read_jsonld: Exception: {}, {}".format(_filename, e)) else: print("Error: metadata file (.jsonld) not found in LiPD archive") except Exception as e: print("Error: Unable to find jsonld file in LiPD archive. This may be a corrupt file.") logger_jsons.error("Error: Unable to find jsonld file in LiPD archive. This may be a corrupt file.") logger_jsons.info("exit read_json_from_file") return _d
Find jsonld file in the cwd (or within a 2 levels below cwd), and load it in. :return dict: Jsonld data
entailment
def read_json_from_file(filename): """ Import the JSON data from target file. :param str filename: Target File :return dict: JSON data """ logger_jsons.info("enter read_json_from_file") d = OrderedDict() try: # Load and decode d = demjson.decode_file(filename, decode_float=float) logger_jsons.info("successful read from json file") except FileNotFoundError: # Didn't find a jsonld file. Maybe it's a json file instead? try: d = demjson.decode_file(os.path.splitext(filename)[0] + '.json', decode_float=float) except FileNotFoundError as e: # No json or jsonld file. Exit print("Error: jsonld file not found: {}".format(filename)) logger_jsons.debug("read_json_from_file: FileNotFound: {}, {}".format(filename, e)) except Exception: print("Error: unable to read jsonld file") if d: d = rm_empty_fields(d) logger_jsons.info("exit read_json_from_file") return d
Import the JSON data from target file. :param str filename: Target File :return dict: JSON data
entailment
def idx_num_to_name(L): """ Switch from index-by-number to index-by-name. :param dict L: Metadata :return dict L: Metadata """ logger_jsons.info("enter idx_num_to_name") try: if "paleoData" in L: L["paleoData"] = _import_data(L["paleoData"], "paleo") if "chronData" in L: L["chronData"] = _import_data(L["chronData"], "chron") except Exception as e: logger_jsons.error("idx_num_to_name: {}".format(e)) print("Error: idx_name_to_num: {}".format(e)) logger_jsons.info("exit idx_num_to_name") return L
Switch from index-by-number to index-by-name. :param dict L: Metadata :return dict L: Metadata
entailment
def _import_data(sections, crumbs): """ Import the section metadata and change it to index-by-name. :param list sections: Metadata :param str pc: paleo or chron :return dict _sections: Metadata """ logger_jsons.info("enter import_data: {}".format(crumbs)) _sections = OrderedDict() try: for _idx, section in enumerate(sections): _tmp = OrderedDict() # Process the paleo measurement table if "measurementTable" in section: _tmp["measurementTable"] = _idx_table_by_name(section["measurementTable"], "{}{}{}".format(crumbs, _idx, "measurement")) # Process the paleo model if "model" in section: _tmp["model"] = _import_model(section["model"], "{}{}{}".format(crumbs, _idx, "model")) # Get the table name from the first measurement table, and use that as the index name for this table _table_name = "{}{}".format(crumbs, _idx) # If we only have generic table names, and one exists already, don't overwrite. Create dynamic name if _table_name in _sections: _table_name = "{}_{}".format(_table_name, _idx) # Put the final product into the output dictionary. Indexed by name _sections[_table_name] = _tmp except Exception as e: logger_jsons.error("import_data: Exception: {}".format(e)) print("Error: import_data: {}".format(e)) logger_jsons.info("exit import_data: {}".format(crumbs)) return _sections
Import the section metadata and change it to index-by-name. :param list sections: Metadata :param str pc: paleo or chron :return dict _sections: Metadata
entailment
def _import_model(models, crumbs): """ Change the nested items of the paleoModel data. Overwrite the data in-place. :param list models: Metadata :param str crumbs: Crumbs :return dict _models: Metadata """ logger_jsons.info("enter import_model".format(crumbs)) _models = OrderedDict() try: for _idx, model in enumerate(models): # Keep the original dictionary, but replace the three main entries below # Do a direct replacement of chronModelTable columns. No table name, no table work needed. if "summaryTable" in model: model["summaryTable"] = _idx_table_by_name(model["summaryTable"], "{}{}{}".format(crumbs, _idx, "summary")) # Do a direct replacement of ensembleTable columns. No table name, no table work needed. if "ensembleTable" in model: model["ensembleTable"] = _idx_table_by_name(model["ensembleTable"], "{}{}{}".format(crumbs, _idx, "ensemble")) if "distributionTable" in model: model["distributionTable"] = _idx_table_by_name(model["distributionTable"], "{}{}{}".format(crumbs, _idx, "distribution")) _table_name = "{}{}".format(crumbs, _idx) _models[_table_name] = model except Exception as e: logger_jsons.error("import_model: {}".format(e)) print("Error: import_model: {}".format(e)) logger_jsons.info("exit import_model: {}".format(crumbs)) return _models
Change the nested items of the paleoModel data. Overwrite the data in-place. :param list models: Metadata :param str crumbs: Crumbs :return dict _models: Metadata
entailment
def _idx_table_by_name(tables, crumbs): """ Import summary, ensemble, or distribution data. :param list tables: Metadata :return dict _tables: Metadata """ _tables = OrderedDict() try: for _idx, _table in enumerate(tables): # Use "name" as tableName _name = "{}{}".format(crumbs, _idx) # Call idx_table_by_name _tmp = _idx_col_by_name(_table) if _name in _tables: _name = "{}_{}".format(_name, _idx) _tmp["tableName"] = _name _tables[_name] = _tmp except Exception as e: logger_jsons.error("idx_table_by_name: {}".format(e)) print("Error: idx_table_by_name: {}".format(e)) return _tables
Import summary, ensemble, or distribution data. :param list tables: Metadata :return dict _tables: Metadata
entailment
def _idx_col_by_name(table): """ Iter over columns list. Turn indexed-by-num list into an indexed-by-name dict. Keys are the variable names. :param dict table: Metadata :return dict _table: Metadata """ _columns = OrderedDict() # Iter for each column in the list try: for _column in table["columns"]: try: _name = _column["variableName"] if _name in _columns: _name = get_appended_name(_name, _columns) _columns[_name] = _column except Exception as e: print("Error: idx_col_by_name: inner: {}".format(e)) logger_jsons.info("idx_col_by_name: inner: {}".format(e)) table["columns"] = _columns except Exception as e: print("Error: idx_col_by_name: {}".format(e)) logger_jsons.error("idx_col_by_name: {}".format(e)) return table
Iter over columns list. Turn indexed-by-num list into an indexed-by-name dict. Keys are the variable names. :param dict table: Metadata :return dict _table: Metadata
entailment
def get_csv_from_json(d): """ Get CSV values when mixed into json data. Pull out the CSV data and put it into a dictionary. :param dict d: JSON with CSV values :return dict: CSV values. (i.e. { CSVFilename1: { Column1: [Values], Column2: [Values] }, CSVFilename2: ... } """ logger_jsons.info("enter get_csv_from_json") csv_data = OrderedDict() if "paleoData" in d: csv_data = _get_csv_from_section(d, "paleoData", csv_data) if "chronData" in d: csv_data = _get_csv_from_section(d, "chronData", csv_data) logger_jsons.info("exit get_csv_from_json") return csv_data
Get CSV values when mixed into json data. Pull out the CSV data and put it into a dictionary. :param dict d: JSON with CSV values :return dict: CSV values. (i.e. { CSVFilename1: { Column1: [Values], Column2: [Values] }, CSVFilename2: ... }
entailment
def _get_csv_from_section(d, pc, csv_data): """ Get csv from paleo and chron sections :param dict d: Metadata :param str pc: Paleo or chron :return dict: running csv data """ logger_jsons.info("enter get_csv_from_section: {}".format(pc)) for table, table_content in d[pc].items(): # Create entry for this table/CSV file (i.e. Asia-1.measTable.PaleoData.csv) # Note: Each table has a respective CSV file. csv_data[table_content['filename']] = OrderedDict() for column, column_content in table_content['columns'].items(): # Set the "values" into csv dictionary in order of column "number" csv_data[table_content['filename']][column_content['number']] = column_content['values'] logger_jsons.info("exit get_csv_from_section: {}".format(pc)) return csv_data
Get csv from paleo and chron sections :param dict d: Metadata :param str pc: Paleo or chron :return dict: running csv data
entailment
def remove_csv_from_json(d): """ Remove all CSV data 'values' entries from paleoData table in the JSON structure. :param dict d: JSON data - old structure :return dict: Metadata dictionary without CSV values """ logger_jsons.info("enter remove_csv_from_json") # Check both sections if "paleoData" in d: d = _remove_csv_from_section(d, "paleoData") if "chronData" in d: d = _remove_csv_from_section(d, "chronData") logger_jsons.info("exit remove_csv_from_json") return d
Remove all CSV data 'values' entries from paleoData table in the JSON structure. :param dict d: JSON data - old structure :return dict: Metadata dictionary without CSV values
entailment
def _remove_csv_from_section(d, pc): """ Remove CSV from metadata in this section :param dict d: Metadata :param str pc: Paleo or chron :return dict: Modified metadata """ logger_jsons.info("enter remove_csv_from_json: {}".format(pc)) for table, table_content in d[pc].items(): for column, column_content in table_content['columns'].items(): try: # try to delete the values key entry del column_content['values'] except KeyError as e: # if the key doesn't exist, keep going logger_jsons.debug("remove_csv_from_json: KeyError: {}, {}".format(pc, e)) logger_jsons.info("exit remove_csv_from_json: {}".format(pc)) return d
Remove CSV from metadata in this section :param dict d: Metadata :param str pc: Paleo or chron :return dict: Modified metadata
entailment