repo_name stringlengths 1 62 | dataset stringclasses 1
value | lang stringclasses 11
values | pr_id int64 1 20.1k | owner stringlengths 2 34 | reviewer stringlengths 2 39 | diff_hunk stringlengths 15 262k | code_review_comment stringlengths 1 99.6k |
|---|---|---|---|---|---|---|---|
lm-human-preference-details | github_2023 | python | 19 | vwxyzjn | lewtun | @@ -510,6 +514,37 @@ def train(args: Args):
)
dataloader = DataLoader(dataset, batch_size=args.ppo.local_batch_size)
policy, optimizer, dataloader = accelerator.prepare(policy, optimizer, dataloader)
+ if args.deepspeed:
+ import deepspeed
+
+ deepspeed_states = AcceleratorState().deepsp... | If I'm not mistaken, these config values are set automatically by the accelerator and don't need to be overridden |
lm-human-preference-details | github_2023 | python | 19 | vwxyzjn | lewtun | @@ -510,6 +514,37 @@ def train(args: Args):
)
dataloader = DataLoader(dataset, batch_size=args.ppo.local_batch_size)
policy, optimizer, dataloader = accelerator.prepare(policy, optimizer, dataloader)
+ if args.deepspeed:
+ import deepspeed
+
+ deepspeed_states = AcceleratorState().deepsp... | I think this flag and the one below are false by default, so probably don't need to be set either |
lm-human-preference-details | github_2023 | python | 19 | vwxyzjn | lewtun | @@ -755,7 +790,8 @@ def train(args: Args):
)
with torch.no_grad():
- writer.add_histogram("ppo/val/ratio_hist", ratio, update)
+ if not args.deepspeed: # for some reason there is a OOM with the `writer.add_histogram`
+ writer.add_histogram("ppo/val/ra... | FYI I was able to train 7B models in TRL with ZeRO-2 and didn't need to remove the histogram. On the other hand that was for sentiment tuning, which is less memory intensive than your application here |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,975 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+
+import einops
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+import optax
+import orbax
... | This should go into the jitted function. |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,975 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+
+import einops
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+import optax
+import orbax
... | All of this should go to the jitted function. Also, do not worry about `del` — if I remember correctly, jax compiler will basically do that for you automatically. |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,975 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+
+import einops
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+import optax
+import orbax
... | Advantages calculation should be jitted as well. You should use `jax.scan` to do the `for t in reversed(range(gen_length)):`
References:
* not directly applicable, but here is Cleanba PPO GAE `jax.lax.scan`: https://github.com/vwxyzjn/cleanba/blob/81dca0054c8c0930046a2d12b07ada2945014d01/cleanba/cleanba_ppo.py#L558... |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,975 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+
+import einops
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+import optax
+import orbax
... | Everything here goes into jitted function |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,975 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+
+import einops
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+import optax
+import orbax
... | jax.lax.scan.See https://github.com/vwxyzjn/cleanba/blob/81dca0054c8c0930046a2d12b07ada2945014d01/cleanba/cleanba_ppo.py#L638-L654 |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,975 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+
+import einops
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+import optax
+import orbax
... | You probably just need a single `pmap` on a `train` function that does all the pmaps you have here together. |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,1024 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+import einops
+from clu import metrics
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+imp... | To simplify, maybe just to clone `ref_policy_params = copy.deepcopy(policy_params)`? |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,1024 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+import einops
+from clu import metrics
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+imp... | Avoid using `jax_utils.replicate` in this case I think. Just pass in regular numpy array. I am worried `jax_utils.replicate` creates jax arrays somehow, which if not done under the jitted function can be very slow. |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,1024 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+import einops
+from clu import metrics
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+imp... | Anytime jax related computation happens outside jitted function it can be slow. Maybe do something similar to https://github.com/vwxyzjn/cleanba/blob/81dca0054c8c0930046a2d12b07ada2945014d01/cleanba/cleanba_ppo.py#L655-L659? |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,1024 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+import einops
+from clu import metrics
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+imp... | This is jax computation happenning outside jitted function, slow. |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,1024 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+import einops
+from clu import metrics
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+imp... | computation happenning outside jitted function |
lm-human-preference-details | github_2023 | python | 18 | vwxyzjn | vwxyzjn | @@ -0,0 +1,1024 @@
+import functools
+import os
+import time
+from dataclasses import asdict, dataclass, field
+from types import SimpleNamespace
+from typing import List, Optional
+import einops
+from clu import metrics
+import flax
+import flax.linen as nn
+import jax
+import jax.numpy as jnp
+import numpy as np
+imp... | learning rate annealing happening outside of `otpax`is a big no no. I remember seeing it makes the training twice as slow. Instead do https://github.com/vwxyzjn/cleanba/blob/81dca0054c8c0930046a2d12b07ada2945014d01/cleanba/cleanba_ppo.py#L497 |
lm-human-preference-details | github_2023 | python | 17 | vwxyzjn | vwxyzjn | @@ -402,11 +401,21 @@ def normalize(
generation_config,
):
# number of minibatches for computing the normalization statistics
- n_batches = ceil_div(args.local_normalize_samples, args.rollout_batch_size)
+ n_batches = ceil_div(args.normalize_samples, args.rollout_batch_size) | This should still be `local_normalize_samples`, right? |
lm-human-preference-details | github_2023 | python | 17 | vwxyzjn | vwxyzjn | @@ -636,7 +649,7 @@ def train(args: Args):
start_text=args.task.start_text,
end_text=args.task.end_text,
)
- normalization_dataloader = DataLoader(normalization_dataset, batch_size=args.rollout_batch_size)
+ normalization_dataloader = DataLoader(normalization_dataset, batch_size=args.rollou... | Imagine we have
* `rollout_batch_size = 16`
* `local_rollout_batch_size = 1`
* `len(devices) = 8` (8 GPUs)
* `world_size=2` (2 distributed training processes, 16 GPUs total)
Then the data loader should have a batch size of 8, which we then shard to 8 devices locally in each of the two processes.
|
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | Maybe just do `optimizer = optax.MultiSteps(optimizer, args.gradient_accumulation_steps)`. This will make logging learning rate easier. If `gradient_accumulation_steps==1`, the behavior is equivalent.
See
https://github.com/vwxyzjn/cleanba/blob/352aa6ff495c7cdca02890ec92feb01fd20c2e63/cleanba/cleanba_impala.py#L53... |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | Maybe use the boilerplate here https://github.com/vwxyzjn/cleanba/blob/352aa6ff495c7cdca02890ec92feb01fd20c2e63/cleanba/cleanba_impala.py#L459-L503, which helps you scale beyond 8 GPU with local_devices and global_devices.
Also e.g.,
```
args.world_size = jax.process_count()
args.local_rank = jax.pro... |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | This is more of a style thing — in CleanRL's style, we try to avoid creating a function unless it is actually used multiple times like `right_padding_to_left_padding`. This is because when reading this line of code I would then need to traverse up to see where the code is defined, causing some jumping around looking ex... |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | This can be changed to `args.seed + jax.process_index() * 100003 # Prime` |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | My ideal format looks like the following, where we eliminate the need for `RewardModelParams`
```
class AutoModelForCausalLMWithRewardHead(nn.Module):
model_name: str
@nn.compact
def __call__(self, x):
backbone = transformers.FlaxAutoModelForCausalLM.from_pretrained(
self.mo... |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... |
Btw optax's adam can be made equivalent to tensorflow's adam with the following optimizer. We should include this `use_tensorflow_adam` in the jax version as well.
https://gist.github.com/vwxyzjn/7005a81ba39deb3bc8043041bd715be1#file-main_jax-py-L69-L78 |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | The common practice seems to be creating an RNG then splitting them in the beginning. E.g.,
```
key = jax.random.PRNGKey(args.seed)
key, network_key, actor_key, critic_key = jax.random.split(key, 4)
``` |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | if args.track and args.local_rank == 0: |
lm-human-preference-details | github_2023 | python | 13 | vwxyzjn | vwxyzjn | @@ -0,0 +1,765 @@
+""" Train jax-based reward model for LM human preference details."""
+import os
+from dataclasses import asdict, dataclass, field
+from typing import Optional
+import time
+import functools
+import numpy as np
+from torch.utils.data import DataLoader, IterableDataset
+import jax
+import jax.numpy as ... | if args.save_path and args.local_rank == 0: |
Queryable | github_2023 | others | 22 | mazzzystar | mazzzystar | @@ -64,7 +64,7 @@ struct SearchResultsView: View {
case .HAS_RESULT:
// Has result
VStack {
- if photoSearcher.totalUnIndexedPhotosNum > 0 {
+ if photoSearcher.totalUnIndexedPhotosNum > 0 || PHPhotoLibrary.authorizationStatus(for: .readWrite) == .limited ... | Reasonable. |
nuxt-open-fetch | github_2023 | typescript | 41 | enkot | enkot | @@ -100,6 +101,11 @@ export default defineNuxtModule<ModuleOptions>({
}
}
+ nuxt.options.alias = {
+ ...nuxt.options.alias,
+ '#nuxt-open-fetch-schemas': join(nuxt.options.buildDir, 'types', moduleName, 'schemas'), | What do you think about '#open-fetch-schemas'? |
nuxt-open-fetch | github_2023 | typescript | 22 | enkot | IlyaSemenov | @@ -0,0 +1,12 @@
+export default defineEventHandler(async (event) => {
+ const { $fetchPets } = useNuxtOpenFetchServer()
+ const data = await $fetchPets ("/pet/{petId}", {
+ path: {
+ petId: 1,
+ },
+ });
+
+ return ({
+ data,
+ })
+}) | Please lint this (get rid of 'no new line at the end of file'). The project uses [editorconfig](https://github.com/enkot/nuxt-open-fetch/blob/18c4ef73a106f95378fd4313ebeb1ef0c25de22a/.editorconfig#L9) but its setup was ignored. |
nuxt-open-fetch | github_2023 | typescript | 22 | enkot | IlyaSemenov | @@ -12,3 +12,4 @@ export default defineNuxtPlugin(() => {
}), {})
}
})
+ | why the stray change? |
nuxt-open-fetch | github_2023 | others | 23 | enkot | skf-funzt | @@ -1,6 +1,19 @@
# Changelog
+## v0.5.0
+
+[compare changes](https://github.com/enkot/nuxt-open-fetch/compare/v0.4.5...v0.5.0)
+
+### 🚀 Enhancements
+
+- Nitro support ([d30287f](https://github.com/enkot/nuxt-open-fetch/commit/d30287f))
+
+### ❤️ Contributors
+
+- Enkot ([@enkot](http://github.com/enkot))
+- Step... | :star_struck: |
nuxt-open-fetch | github_2023 | others | 17 | enkot | IlyaSemenov | @@ -1 +1,2 @@
typescript.includeWorkspace=true
+imports.autoImport=false | The project has `.editorconfig` which explicitly forces final new line:
https://github.com/enkot/nuxt-open-fetch/blob/55a619189adde50a418c64d1d1da1bd93c0bdc9c/.editorconfig#L9
Please lint your code accordingly. |
magic-spreadsheets.github.io | github_2023 | typescript | 43 | magic-spreadsheets | hihumikan | @@ -2,6 +2,6 @@ import type { ContactDetails } from './src/types';
// お問い合わせ先
export const contactDetails: ContactDetails = {
- text: '@keigomichi',
- href: 'https://twitter.com/keigomichi',
+ text: '@hihumikan', | ```suggestion
text: '@mikan_54951',
``` |
magic-spreadsheets.github.io | github_2023 | typescript | 43 | magic-spreadsheets | hihumikan | @@ -1,6 +1,12 @@
import type { Maintainers } from './src/types';
export const maintainers: Maintainers = {
+ 2024: [
+ {
+ text: '@hihumikan', | ```suggestion
text: '@mikan_54951',
``` |
magic-spreadsheets.github.io | github_2023 | others | 43 | magic-spreadsheets | hihumikan | @@ -33,7 +33,7 @@ const hashtags = encodeURIComponent(`魔法のスプレッドシート,インタ
<meta property="og:type" content="website" />
<meta property="og:image" content="https://magic-spreadsheets.pages.dev/thumbnail.png" />
<meta property="twitter:card" content="summary_large_image" />
- <meta property="twitter:site" ... | ```suggestion
<meta property="twitter:site" content="@mikan_54951" />
``` |
magic-spreadsheets.github.io | github_2023 | others | 43 | magic-spreadsheets | hihumikan | @@ -1,6 +1,6 @@
MIT License
-Copyright (c) 2023 keigomichi
+Copyright (c) 2023 hihumikan | ```suggestion
Copyright (c) 2023 magic-spreadsheets
``` |
magic-spreadsheets.github.io | github_2023 | others | 2 | magic-spreadsheets | keigomichi | @@ -1,15 +1,271 @@
---
---
-<html lang="en">
+<html lang="ja">
<head>
<meta charset="utf-8" />
<link rel="icon" type="image/svg+xml" href="/favicon.svg" />
<meta name="viewport" content="width=device-width" />
<meta name="generator" content={Astro.generator} />
- <title>Astro</title>
+ ... | @cardseditor フォームについて、tally以外で作成することも検討するので、一旦iframeごと削除してもらっていいですか？ |
tonlib-rs | github_2023 | others | 143 | ston-fi | ruslanracheev | @@ -85,4 +85,7 @@
* Impl #BE-2088: Wallet v5 message building
### v0.24.2
* Bump tonlib-sys to 2025.2.2
-
+### v0.24.3
+* Impl #ni: tonaddres::from_msg_address | typo `tonaddress` |
tonlib-rs | github_2023 | others | 115 | ston-fi | dbaranovstonfi | @@ -50,7 +50,7 @@ tokio-retry = "0.3"
tokio-test = "0.4"
ton_liteapi = "0.1.0"
adnl = "2.0"
-tonlib-sys = "=2024.9.0"
+tonlib-sys = "=2024.11.0" | The version of tonlib-sys will be 2024.10.1. |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -329,22 +329,22 @@ mod tests {
#[ignore]
#[test]
fn check_code_hash() -> Result<(), TonCellError> {
- let raw = include_str!("../../resources/wallet/wallet_v3_code.hex");
- let boc = BagOfCells::parse(&hex::decode(raw).map_boc_deserialization_error()?)?;
+ let raw = include_str!("... | I would suggest to use function `BagOfCells::parse_base64` function here and in all similar places below.
|
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -42,19 +108,44 @@ impl WalletVersion {
&self,
workchain: i32,
key_pair: &KeyPair,
+ sub_wallet_id: Option<i32>,
) -> Result<BagOfCells, TonCellError> {
- let mut data_builder = CellBuilder::new();
- data_builder
- .store_u32(32, 0)?
- // seqn... | I think that using `WalletVersion::V1R1` instead of `Self::V1R1` would improve readability a lot. |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -42,19 +108,44 @@ impl WalletVersion {
&self,
workchain: i32,
key_pair: &KeyPair,
+ sub_wallet_id: Option<i32>,
) -> Result<BagOfCells, TonCellError> {
- let mut data_builder = CellBuilder::new();
- data_builder
- .store_u32(32, 0)?
- // seqn... | It is better to explicitly mention all possible WalletVersion members in this match . |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -42,19 +108,44 @@ impl WalletVersion {
&self,
workchain: i32,
key_pair: &KeyPair,
+ sub_wallet_id: Option<i32>,
) -> Result<BagOfCells, TonCellError> {
- let mut data_builder = CellBuilder::new();
- data_builder
- .store_u32(32, 0)?
- // seqn... | We prefer to avoid usage of `unimplemented!` and other code that panics. |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -70,6 +161,150 @@ impl WalletVersion {
}
}
+/// WalletVersion::V1R1 | WalletVersion::V1R2 | WalletVersion::V1R3 | WalletVersion::V2R1 | WalletVersion::V2R2
+pub struct DataV1R1 { | I would suggest to move this struct and the traits to wallet/types.rs |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -70,6 +161,150 @@ impl WalletVersion {
}
}
+/// WalletVersion::V1R1 | WalletVersion::V1R2 | WalletVersion::V1R3 | WalletVersion::V2R1 | WalletVersion::V2R2
+pub struct DataV1R1 {
+ pub seqno: u32,
+ pub public_key: [u8; 32],
+}
+
+impl TryFrom<&Cell> for DataV1R1 {
+ type Error = TonCellError;
+
+ ... | Looks like ` wallet_id` should be i32 to avoid unnecessary typecasts above. However I have noticed that we do not have a corresponding loader function.
Please feel free to add
`pub fn load_i32(&mut self, bit_len: usize) -> Result<i32, TonCellError>` to `pub struct CellParser<'a>` |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -42,19 +108,44 @@ impl WalletVersion {
&self,
workchain: i32,
key_pair: &KeyPair,
+ sub_wallet_id: Option<i32>,
) -> Result<BagOfCells, TonCellError> {
- let mut data_builder = CellBuilder::new();
- data_builder
- .store_u32(32, 0)?
- // seqn... | We avoid using `expect` as it panics. Returning ` TonCellError::InternalError("Invalid public key size".to_string())` looks more suitable. |
tonlib-rs | github_2023 | others | 39 | ston-fi | dbaranovstonfi | @@ -70,6 +161,150 @@ impl WalletVersion {
}
}
+/// WalletVersion::V1R1 | WalletVersion::V1R2 | WalletVersion::V1R3 | WalletVersion::V2R1 | WalletVersion::V2R2
+pub struct DataV1R1 {
+ pub seqno: u32,
+ pub public_key: [u8; 32],
+}
+
+impl TryFrom<&Cell> for DataV1R1 {
+ type Error = TonCellError;
+
+ ... | I would suggest to follow guideline in rust [documentation](https://doc.rust-lang.org/std/convert/trait.TryInto.html) , which recommends to use ` impl From` and `impl TryFrom` instead of `impl Into` and `impl TryInto`respectively.
|
tonlib-rs | github_2023 | others | 40 | ston-fi | ruslanracheev | @@ -1,6 +1,6 @@
[package]
name = "tonlib"
-version = "0.12.1-dev"
+version = "0.12.2" | it is not a minor update. there are incompatible changes. |
tonlib-rs | github_2023 | others | 16 | ston-fi | dbaranovstonfi | @@ -0,0 +1,18 @@
+use tonlib::cell::BagOfCells; | I would recommend to move this test to client_test.rs to preserve consistency |
tonlib-rs | github_2023 | others | 16 | ston-fi | dbaranovstonfi | @@ -65,6 +66,9 @@ pub enum TonResult {
#[serde(rename = "blocks.header")]
BlocksHeader(BlocksHeader),
// tonlib_api.tl, line 228
+ #[serde(rename = "configInfo")]
+ ChainConfigInfo(ChainConfigInfo), | I would suggest preserving the original naming from tonlob_api.tl: ConfigInfo |
tonlib-rs | github_2023 | others | 15 | ston-fi | ruslanracheev | @@ -42,6 +43,8 @@ pub enum TonResult {
// tonlib_api.tl, line 181
#[serde(rename = "smc.runResult")]
SmcRunResult(SmcRunResult),
+ #[serde(rename = "tvm.cell")] | здесь надо указать в комменте на какой строке это в tonlib_api.tl |
tonlib-rs | github_2023 | others | 6 | ston-fi | Vasilisck | @@ -0,0 +1,141 @@
+{ | Видимо гит игнор надо обновить |
tonlib-rs | github_2023 | others | 6 | ston-fi | Vasilisck | @@ -226,6 +226,15 @@ pub trait TonFunctions {
}
}
+ async fn smc_forget(&self, id: i64) -> anyhow::Result<(TonConnection, i64)> {
+ let func = TonFunction::SmcForget { id };
+ let (conn, result) = self.invoke_on_connection(&func).await?; | тут точно надо invoke_on_connection а не invoke ? |
tonlib-rs | github_2023 | others | 6 | ston-fi | Vasilisck | @@ -113,3 +113,33 @@ impl TonContract {
}
}
}
+
+impl Drop for TonContract {
+ fn drop(&mut self) {
+ let runtime = match tokio::runtime::Builder::new_current_thread() | Я думаю стоит переписать на tokio::spawn т.к. в других местах мы его используем. |
tonlib-rs | github_2023 | others | 6 | ston-fi | ruslanracheev | @@ -112,4 +112,23 @@ impl TonContract {
)),
}
}
+
+ async fn forget_state(&self) -> anyhow::Result<()> {
+ if let Ok(state) = self.load_state().await { | зачем мы в методе forget_state сначала его загружаем? |
halo-theme-chirpy | github_2023 | others | 90 | AirboZH | AirboZH | @@ -0,0 +1,22 @@
+<!DOCTYPE html>
+<html
+ xmlns:th="https://www.thymeleaf.org"
+ th:replace="~{modules/layout :: html(title = ${error.status} + ': ' + ${#strings.defaultString(error.title, 'Internal server error')} + ' - ' +${site.title},
+ content = ~{modules/page :: page(~{::content}, '404')})}" | 页面layout可以写为error |
halo-theme-chirpy | github_2023 | others | 90 | AirboZH | AirboZH | @@ -0,0 +1,22 @@
+<!DOCTYPE html>
+<html
+ xmlns:th="https://www.thymeleaf.org"
+ th:replace="~{modules/layout :: html(title = ${error.status} + ': ' + ${#strings.defaultString(error.title, 'Internal server error')} + ' - ' +${site.title},
+ content = ~{modules/page :: page(~{::content}, '404')})}"
+>
+ <th:block ... | 可否根据error.status做出调整，或者也可以做成，`我们遇到一点小错误`之类笼统的说法。
目前这个404的提示可能不适用于其他error |
halo-theme-chirpy | github_2023 | others | 82 | AirboZH | AirboZH | @@ -0,0 +1,46 @@
+<!DOCTYPE html>
+<html lang="en" xmlns:th="http://www.thymeleaf.org">
+
+<th:block th:fragment="relatedPosts">
+ <!-- 阅读建议 -->
+ <aside id="related-posts" aria-labelledby="related-label">
+ <h3 class="mb-4">Further Reading</h3>
+ <div th:if="${not #lists.isEmpty(post.categories)}">
+ <nav... | ```suggestion
<p th:text="${recommendPost.status.excerpt}"></p>
```
|
halo-theme-chirpy | github_2023 | others | 82 | AirboZH | AirboZH | @@ -0,0 +1,44 @@
+<!DOCTYPE html>
+<html lang="en" xmlns:th="http://www.thymeleaf.org">
+
+<th:block th:fragment="relatedPosts">
+ <!-- 阅读建议 -->
+ <aside id="related-posts" aria-labelledby="related-label">
+ <h3 class="mb-4">Further Reading</h3>
+ <div th:if="${not #lists.isEmpty(post.categories)}">
+ <nav... | 可以判断`postCursor.next`是否为空
为空则为
```html
<div
class="btn btn-outline-primary disabled"
prompt="{{ site.data.locales[include.lang].post.button.previous }}"
>
<p>-</p>
</div>
```
参考https://github.com/AirboZH/halo-theme-chirpy/blob/master/src/_layouts/_includes/post-nav.html |
halo-theme-chirpy | github_2023 | others | 58 | AirboZH | AirboZH | @@ -3,10 +3,18 @@
<footer th:fragment="footer()">
<div class="container px-lg-4">
<div
- class="d-flex justify-content-center align-items-center text-muted mx-md-3"
+ class=" d-flex justify-content-center text-muted flex-lg-row justify-content-lg-between align-items-lg-center pb-lg-3 break-... | 希望在这有一个备案配置value为空的判断。不然如果这个信息为空，页脚文字都是靠右的
当前效果：
删除后效果：
|
awesome-ai-safety | github_2023 | others | 4 | Giskard-AI | alexcombessie | @@ -48,17 +50,23 @@ You can browse papers by Machine Learning task category, and use hashtags like `
* [Anchors: High-Precision Model-Agnostic Explanations](https://homes.cs.washington.edu/~marcotcr/aaai18.pdf) (Ribeiro et al., 2018) `#Explainability`
* [Explanation-Based Human Debugging of NLP Models: A Survey](http... | ```suggestion
* [Diversity in Faces](https://arxiv.org/abs/1901.10436) (Merler et al.) `#Fairness` `#Accuracy`
``` |
ngxtension-platform | github_2023 | typescript | 587 | ngxtension | nartc | @@ -0,0 +1,18 @@
+import { HttpContext } from '@angular/common/http';
+
+/**
+ * Merge multiple HttpContext.
+ *
+ * @param contexts Two or more http contexts to be merged.
+ * @returns A merged HttpContext.
+ *
+ */
+
+export function mergeHttpContext(...contexts: HttpContext[]): HttpContext {
+ return contexts.reduce... | nit/question: I'm not too familiar with `HttpContext` so I might miss something but why can't we use the first context here instead of a new `HttpContext` instance? |
ngxtension-platform | github_2023 | others | 450 | ngxtension | eneajaho | @@ -50,11 +65,52 @@ class TestComponent {
}
```
-Or, if we want to transform the params, we can pass a function to `injectParams`.
+#### Transform
+
+If you want to additional parse the specific param, we can pass a `parse` function. | ```suggestion
If you want to parse the specific param, you can pass a `parse` function.
``` |
ngxtension-platform | github_2023 | others | 450 | ngxtension | eneajaho | @@ -50,11 +65,52 @@ class TestComponent {
}
```
-Or, if we want to transform the params, we can pass a function to `injectParams`.
+#### Transform
+
+If you want to additional parse the specific param, we can pass a `parse` function.
```ts
-@Component()
+@Component({
+ template: `
+ @if (user(); as user) {
+ ... | Should we update this to use rxResource? At leats that's how I'd use it nowdays. What do you think? |
ngxtension-platform | github_2023 | typescript | 450 | ngxtension | eneajaho | @@ -1,31 +1,44 @@
-import { assertInInjectionContext, inject, type Signal } from '@angular/core';
+import { inject, type Signal } from '@angular/core';
import { toSignal } from '@angular/core/rxjs-interop';
import { ActivatedRoute, type Params } from '@angular/router';
+import { assertInjector } from 'ngxtension/asse... | We can write a migration schematic when we remove this one. So it's smooth for everyone when they update. |
ngxtension-platform | github_2023 | others | 528 | ngxtension | eneajaho | @@ -0,0 +1,95 @@
+---
+title: Queries Migration
+description: Schematics for migrating from decorator-based Queries to Signal-based Queries
+entryPoint: plugin/src/generators/convert-queries
+badge: stable
+contributors: ['enea-jahollari']
+--- | This needs to be updated |
ngxtension-platform | github_2023 | others | 528 | ngxtension | eneajaho | @@ -0,0 +1,95 @@
+---
+title: Queries Migration
+description: Schematics for migrating from decorator-based Queries to Signal-based Queries
+entryPoint: plugin/src/generators/convert-queries
+badge: stable
+contributors: ['enea-jahollari']
+---
+
+Recent releases of Angular have deprecated the `@HostBinding` and `@Host... | The decorators are not yet deprecated |
ngxtension-platform | github_2023 | typescript | 528 | ngxtension | eneajaho | @@ -0,0 +1,108 @@
+import { Tree } from '@nx/devkit';
+import { createTreeWithEmptyWorkspace } from '@nx/devkit/testing';
+
+import { convertHostBindingGenerator } from './generator';
+import { ConvertHostBindingGeneratorSchema } from './schema';
+
+const filesMap = { | I'd like to see some more test cases. And before we merge this one, we will need to make sure that we don't break apps. |
ngxtension-platform | github_2023 | typescript | 543 | ngxtension | michael-small | @@ -0,0 +1,184 @@
+import {
+ computed,
+ Injector,
+ Signal,
+ signal,
+ WritableSignal,
+} from '@angular/core';
+import { explicitEffect } from 'ngxtension/explicit-effect';
+
+interface SignalHistoryRecord<T> {
+ value: T;
+ timestamp: number;
+}
+
+/**
+ * Creates a history record with the current timestamp.
+ * @... | ```suggestion
* @default 100
``` |
ngxtension-platform | github_2023 | typescript | 543 | ngxtension | JeanMeche | @@ -0,0 +1,184 @@
+import {
+ computed,
+ Injector,
+ Signal,
+ signal,
+ WritableSignal,
+} from '@angular/core';
+import { explicitEffect } from 'ngxtension/explicit-effect';
+
+interface SignalHistoryRecord<T> {
+ value: T;
+ timestamp: number;
+}
+
+/**
+ * Creates a history record with the current timestamp.
+ * @... | I personnaly don't like the "inject" part of the name.
You don't inject a signal, you create one, hense I would either go with `signalHistory()` or `createSignalHistory()` |
ngxtension-platform | github_2023 | typescript | 543 | ngxtension | JeanMeche | @@ -0,0 +1,184 @@
+import {
+ computed,
+ Injector,
+ Signal,
+ signal,
+ WritableSignal,
+} from '@angular/core';
+import { explicitEffect } from 'ngxtension/explicit-effect';
+
+interface SignalHistoryRecord<T> {
+ value: T;
+ timestamp: number;
+}
+
+/**
+ * Creates a history record with the current timestamp.
+ * @... | I'd guard that function with `options?.injector && assertInInjectionContext(...);` |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | JeanMeche | @@ -0,0 +1,134 @@
+import { JsonPipe } from '@angular/common';
+import { Component, effect, inject } from '@angular/core';
+import { FormsModule } from '@angular/forms';
+import { Router } from '@angular/router';
+import { linkedQueryParam } from 'ngxtension/linked-query-param';
+
+@Component({
+ standalone: true,
+ te... | You have a todo left here |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | JeanMeche | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | ```suggestion
private _router = inject(Router);
```
For consistency ? |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | JeanMeche | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | ```suggestion
private _schedulerNotifier = createNotifier();
``` |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | JeanMeche | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | if `config` is nullable, it should be reflected on the param type. |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | JeanMeche | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | nit: I'd drop the curly only for the one liners |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,72 @@
+import { JsonPipe } from '@angular/common';
+import { Component, effect, inject, untracked } from '@angular/core';
+import { FormsModule } from '@angular/forms';
+import { Title } from '@angular/platform-browser';
+import { linkedQueryParam } from 'ngxtension/linked-query-param';
+
+@Component({
+ sta... | question: intended comment? |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,72 @@
+import { JsonPipe } from '@angular/common';
+import { Component, effect, inject, untracked } from '@angular/core';
+import { FormsModule } from '@angular/forms';
+import { Title } from '@angular/platform-browser';
+import { linkedQueryParam } from 'ngxtension/linked-query-param';
+
+@Component({
+ sta... | nit: this seems a bit confusing to me. The `effect` seems to rely on `searchQuery` to update the `title` but the actual `searchQuery` value is in `untracked`? I am aware that the track of `searchQuery()` happens in the `console.log` |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,134 @@
+import { JsonPipe } from '@angular/common';
+import { Component, effect, inject } from '@angular/core';
+import { FormsModule } from '@angular/forms';
+import { Router } from '@angular/router';
+import { linkedQueryParam } from 'ngxtension/linked-query-param';
+
+@Component({
+ standalone: true,
+ te... | nit: I wouldn't spawn an `effect` for each `filterState` key |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | nit: my personal style is that if `then()` call is empty, I'd use `void` to annotate that: I skip the return value of this expression call.
```suggestion
untracked(() => void this.navigate());
``` |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | nit: if `config` isn't required as inferred by `config ?? {}` below, provide the default value to the parameter
```suggestion
setCurrentNavigationExtras(config: Partial<NavigateMethodFields> = {}) {
``` |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | issue: I see this as an issue because of the semantic of Serialization in general. Pairing `parse` with `serialize` seems odd to me. I suggest `deserialize` and `serialize` even though `deserialize` is longer than `parse`, the semantic is clear as to what Serialization means in this case for query params. |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | nit/suggestion: I would make it explicitly clear that we'll be overriding `set` and `update` to allow `null`
```suggestion
): Omit<WritableSignal<T>, 'set' | 'update'> & {
set: SignalSetFn<T | null>;
update: SignalUpdateFn<T | null>;
};
``` |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | nit: I would move this error case to the top of the function even before asserting the Injector |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | question: would `export function paramToNumber(): number | null` work? It looks like as soon as I provide a `config` object, the first overload takes over and I can't provide `null | undefined` to `defaultValue` anyway so why would the second overload needs to declare the optional `config?` object at all? |
ngxtension-platform | github_2023 | typescript | 526 | ngxtension | nartc | @@ -0,0 +1,414 @@
+import {
+ effect,
+ inject,
+ Injectable,
+ Injector,
+ runInInjectionContext,
+ signal,
+ untracked,
+ WritableSignal,
+} from '@angular/core';
+import { toSignal } from '@angular/core/rxjs-interop';
+import {
+ ActivatedRoute,
+ NavigationExtras,
+ Params,
+ Router,
+} from '@angular/router';
+imp... | question: same as above about `paramToNumber` overload |
ngxtension-platform | github_2023 | typescript | 525 | ngxtension | eneajaho | @@ -124,29 +124,69 @@ const internalInjectLocalStorage = <R>(
const localStorage = inject(NGXTENSION_LOCAL_STORAGE);
const destroyRef = inject(DestroyRef);
- const initialStoredValue = goodTry(() => localStorage.getItem(key));
- const initialValue = initialStoredValue
- ? (goodTry(() => parse(initialStoredV... | This looks like a job for observables to be honest. I'd like to see this part converted to a subject that emits and this way we can handle things easily. We have too many effects at this point inside the utility. We need only one I guess, the others can be replaced by the subject subscription. What do you think? |
ngxtension-platform | github_2023 | typescript | 525 | ngxtension | nartc | @@ -124,27 +123,74 @@ const internalInjectLocalStorage = <R>(
const localStorage = inject(NGXTENSION_LOCAL_STORAGE);
const destroyRef = inject(DestroyRef);
- const initialStoredValue = goodTry(() => localStorage.getItem(key));
- const initialValue = initialStoredValue
- ? (goodTry(() => parse(initialStoredV... | nit: use DI for `window` |
ngxtension-platform | github_2023 | typescript | 487 | ngxtension | eneajaho | @@ -271,9 +272,7 @@ export function derivedAsync<T>(
untracked(() => sourceEvent$.next(newSource));
} else {
// if the new source is not an observable or a promise, we set the value immediately
- untracked(() =>
- sourceValue.set({ kind: StateKind.Value, value: newSource as T }),
- );
+ untra... | Hi @MillerSvt
We can refactor it a bit more now to just be:
```ts
// we untrack the source$.next() so that we don't register other signals as dependencies
untracked(() => {
sourceEvent$.next(
isObservable(newSource) || isPromise(newSource) ? newSource : of(newSource as T)
);
});
``` |
ngxtension-platform | github_2023 | typescript | 508 | ngxtension | eneajaho | @@ -29,6 +29,16 @@ describe('injectLocalStorage', () => {
});
}));
+ it('should return data of defaultValue', () => {
+ TestBed.runInInjectionContext(() => {
+ const defaultValue = 'default';
+ const localStorageSignal = injectLocalStorage<string>(key, { | Can we include a test that only checks for the type and not just the value? |
ngxtension-platform | github_2023 | others | 508 | ngxtension | eneajaho | @@ -55,9 +55,8 @@ Options to configure the behavior of the local storage signal.
Here's a basic example of using `injectLocalStorage`:
```typescript
-const username = injectLocalStorage<string>('username', {
- defaultValue: 'Anonymous',
- storageSync: true,
+const username = injectLocalStorage<string | undefined>('... | I would move this below to showcase the storageSync method, instead of showcasing it as a basic example.
A basic example would just be : `const username = injectLocalStorage<string>('username');` and tell the developer that it will also contain the null or undefined (whatever we use as default) value in the type |
ngxtension-platform | github_2023 | typescript | 282 | ngxtension | JeanMeche | @@ -0,0 +1,285 @@
+import {
+ Injector,
+ computed,
+ effect,
+ isSignal,
+ signal,
+ untracked,
+ type Signal,
+} from '@angular/core';
+import { assertInjector } from 'ngxtension/assert-injector';
+import { computedPrevious } from 'ngxtension/computed-previous';
+
+interface CreateResourceOptions {
+ injector?: Injec... | ```suggestion
if (isPromise(p)) {
``` |
ngxtension-platform | github_2023 | typescript | 282 | ngxtension | JeanMeche | @@ -0,0 +1,285 @@
+import {
+ Injector,
+ computed,
+ effect,
+ isSignal,
+ signal,
+ untracked,
+ type Signal,
+} from '@angular/core';
+import { assertInjector } from 'ngxtension/assert-injector';
+import { computedPrevious } from 'ngxtension/computed-previous';
+
+interface CreateResourceOptions {
+ injector?: Injec... | ```suggestion
state.set(trigger() !== previousTrigger() ? 'refreshing' : 'pending');
``` |
ngxtension-platform | github_2023 | typescript | 282 | ngxtension | ilirbeqirii | @@ -0,0 +1,285 @@
+import {
+ Injector,
+ computed,
+ effect,
+ isSignal,
+ signal,
+ untracked,
+ type Signal,
+} from '@angular/core';
+import { assertInjector } from 'ngxtension/assert-injector';
+import { computedPrevious } from 'ngxtension/computed-previous';
+
+interface CreateResourceOptions {
+ injector?: Injec... | maybe this can be renamed to 'status' thus matching the Type Parameter itself too? |
ngxtension-platform | github_2023 | others | 448 | ngxtension | nartc | @@ -5,8 +5,8 @@ import { Icon } from '@astrojs/starlight/components';
const badge = Astro.props.entry.data.badge;
const entryPoint = Astro.props.entry.data.entryPoint;
-const bundleJsUrl = `https://deno.bundlejs.com?q=ngxtension/${entryPoint}&treeshake=[*]&config={%22esbuild%22:{%22external%22:[%22rxjs%22,%22@angul... | ```suggestion
const sourceCodeUrl = `https://github.com/ngxtension/ngxtension-platform/tree/main/libs/${entryPoint}`;
``` |
ngxtension-platform | github_2023 | typescript | 467 | ngxtension | ajitzero | @@ -138,11 +138,6 @@ describe(signalSlice.name, () => {
state.age();
expect(testFn).toHaveBeenCalled();
});
-
- it('should connect lazy source after signal value is accessed', () => {
- state().age; | What is this a duplicate of?
The previous one is `state.age()` while this one is `state().age` |
ngxtension-platform | github_2023 | others | 420 | ngxtension | michael-small | @@ -0,0 +1,57 @@
+---
+title: Convert to SFC components migration
+description: Schematics for converting Angular components to SFC components
+entryPoint: convert-to-sfc
+badge: stable
+contributors: ['enea-jahollari']
+---
+
+Angular components can have inline templates or have a separate template file. The inline te... | extra ```bash that throws off formatting |
ngxtension-platform | github_2023 | typescript | 420 | ngxtension | ilirbeqirii | @@ -0,0 +1,171 @@
+import {
+ formatFiles,
+ getProjects,
+ joinPathFragments,
+ logger,
+ readJson,
+ readProjectConfiguration,
+ Tree,
+ visitNotIgnoredFiles,
+} from '@nx/devkit';
+import { readFileSync } from 'node:fs';
+import { dirname } from 'node:path';
+import { exit } from 'node:process';
+import { Node, Synt... | Is it possible to create a reusable function for preparing "templatePath", and then reading its content? I also noticed a similar code in the "self-closing tags" schematic. |
ngxtension-platform | github_2023 | typescript | 361 | ngxtension | nartc | @@ -157,16 +131,7 @@ export function signalSlice<
effects?: ( | nit: we can use `/** @deprecated */` JSDoc here to make the intention of deprecating `effects` clearer |
ngxtension-platform | github_2023 | others | 404 | ngxtension | ajitzero | @@ -0,0 +1,6 @@
+{
+ "name": "Fabien Dehopré",
+ "twitter": "https://twitter.com/FabienDehopre",
+ "github": "https://github.com/FabienDehopre",
+ "linkedin": "https://www.linkedin.com/in/fabien1979/?lipi=urn%3Ali%3Apage%3Ad_flagship3_feed%3BbRilI4BMRLygBScKCu5i%2Fw%3D%3D" | minor: 😅
```suggestion
"linkedin": "https://www.linkedin.com/in/fabien1979/"
``` |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.