content-core 1.2.1__py3-none-any.whl → 1.2.2__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.

This version of content-core might be problematic. Click here for more details.

Files changed (7) hide show
  1. content_core/processors/url.py +2 -2
  2. content_core/processors/youtube.py +4 -3
  3. {content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/METADATA +1 -1
  4. {content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/RECORD +7 -7
  5. {content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/WHEEL +0 -0
  6. {content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/entry_points.txt +0 -0
  7. {content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/licenses/LICENSE +0 -0
content_core/processors/url.py CHANGED
@@ -38,10 +38,10 @@ async def url_provider(state: ProcessSourceState):
38
38
  or mime in SUPPORTED_FITZ_TYPES
39
39
  or mime in SUPPORTED_OFFICE_TYPES
40
40
  ):
41
- logger.warning(f"Identified type for {url}: {mime}")
41
+ logger.debug(f"Identified type for {url}: {mime}")
42
42
  return_dict["identified_type"] = mime
43
43
  else:
44
- logger.warning(f"Identified type for {url}: article")
44
+ logger.debug(f"Identified type for {url}: article")
45
45
  return_dict["identified_type"] = "article"
46
46
  return return_dict
47
47
 
content_core/processors/youtube.py CHANGED
@@ -3,12 +3,13 @@ import ssl
3
3
 
4
4
  import aiohttp
5
5
  from bs4 import BeautifulSoup
6
+ from youtube_transcript_api import YouTubeTranscriptApi # type: ignore
7
+ from youtube_transcript_api.formatters import TextFormatter # type: ignore
8
+
6
9
  from content_core.common import ProcessSourceState
7
10
  from content_core.common.exceptions import NoTranscriptFound
8
11
  from content_core.config import CONFIG
9
12
  from content_core.logging import logger
10
- from youtube_transcript_api import YouTubeTranscriptApi # type: ignore
11
- from youtube_transcript_api.formatters import TextFormatter # type: ignore
12
13
 
13
14
  ssl._create_default_https_context = ssl._create_unverified_context
14
15
 
@@ -172,7 +173,7 @@ async def extract_youtube_transcript(state: ProcessSourceState):
172
173
  """
173
174
 
174
175
  assert state.url, "No URL provided"
175
- logger.warning(f"Extracting transcript from URL: {state.url}")
176
+ logger.debug(f"Extracting transcript from URL: {state.url}")
176
177
  languages = CONFIG.get("youtube_transcripts", {}).get(
177
178
  "preferred_languages", ["en", "es", "pt"]
178
179
  )
{content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/METADATA RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: content-core
3
- Version: 1.2.1
3
+ Version: 1.2.2
4
4
  Summary: Extract what matters from any media source. Available as Python Library, macOS Service, CLI and MCP Server
5
5
  Author-email: LUIS NOVO <lfnovo@gmail.com>
6
6
  License-File: LICENSE
{content_core-1.2.1.dist-info → content_core-1.2.2.dist-info}/RECORD RENAMED
@@ -28,15 +28,15 @@ content_core/processors/docling.py,sha256=lf_NHh255gn4d2EymJYqyH2QiAgQDiJCY3t6Ne
28
28
  content_core/processors/office.py,sha256=DXkfmjqUhmhP6rJaO5Z5Y9sv-iK0zaPZ3waynFIPtsk,12153
29
29
  content_core/processors/pdf.py,sha256=TTDhfV2INtXumFDjLJFNMRfpbJ_tqwIcSBDzuThKxJI,10617
30
30
  content_core/processors/text.py,sha256=kKHA60-NYjLmCTYUnk8TdJxQQ0Shkg-K61Ezqaelz7k,1158
31
- content_core/processors/url.py,sha256=YoWw2CjZbqSKBi1CpY0Qowu4hfqGVGJjLZEXUjz7wxs,7536
31
+ content_core/processors/url.py,sha256=To0LTtMVNN3M83CdodQaZFuU7-IMM5w9QOHRKNV8PVI,7532
32
32
  content_core/processors/video.py,sha256=3WnZwTswvTLm8PtQhKwoqJ2BH6YZi62dMUjALwJiebo,5196
33
- content_core/processors/youtube.py,sha256=MOeZboVfM9_C87L5mnUVvsbQeKoznwJoYn1wP1_hA_U,7869
33
+ content_core/processors/youtube.py,sha256=_qvxI9qTdxu3l1fKLuJARFt8KtZVFJ3JJBLkq1hAAXo,7868
34
34
  content_core/tools/__init__.py,sha256=DuJmd7fE-NpDvLP8IW1XY5MUkAQcdks52rn2jk4N8jQ,231
35
35
  content_core/tools/cleanup.py,sha256=5IdKedsFyRQMdYzgFSKtsfyxJldbroXQXHesHICNENI,523
36
36
  content_core/tools/extract.py,sha256=-r2_jsuMMXyXxGVqWhh1ilNPo_UMYAbw3Pkp1FzPy5g,577
37
37
  content_core/tools/summarize.py,sha256=DPfeglLWB08q8SvHrsKpOKZ35XjduUDs2J02ISwjdj0,596
38
- content_core-1.2.1.dist-info/METADATA,sha256=1LpANnMvECxIekt6kKQr0hnZ1ULGaD2xEmhRh_uzTdk,19676
39
- content_core-1.2.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
40
- content_core-1.2.1.dist-info/entry_points.txt,sha256=ifbBxw37b7gAxZXoduS15KtqHuMHuU58STRkEmgM2zA,147
41
- content_core-1.2.1.dist-info/licenses/LICENSE,sha256=myj0z2T4qIkenCgLsRfx7Wk6UqCQNj5c7O14Qx4zpGg,1066
42
- content_core-1.2.1.dist-info/RECORD,,
38
+ content_core-1.2.2.dist-info/METADATA,sha256=876oFuYDvAzsfOBRRe9t04SprQZne0Xg4BIGhOnzshs,19676
39
+ content_core-1.2.2.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
40
+ content_core-1.2.2.dist-info/entry_points.txt,sha256=ifbBxw37b7gAxZXoduS15KtqHuMHuU58STRkEmgM2zA,147
41
+ content_core-1.2.2.dist-info/licenses/LICENSE,sha256=myj0z2T4qIkenCgLsRfx7Wk6UqCQNj5c7O14Qx4zpGg,1066
42
+ content_core-1.2.2.dist-info/RECORD,,