github/gallery-dl
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[foolfuuka] match 4chan filenames (#2577)

Browse Source 
introduce two new metadata fields:
- filename_media: original filename of file uploaded to 4chan
- timestamp_ms  : timestamp with millisecond precision (tim)
This commit is contained in:
Mike Fährmann
2022-05-14 14:18:50 +02:00
parent 6ca998fc2b
commit bd6ec5c352
1 changed files with 5 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
gallery_dl/extractor/foolfuuka.py
View File

@@ -16,6 +16,7 @@ import itertools
class FoolfuukaExtractor(BaseExtractor): class FoolfuukaExtractor(BaseExtractor):
"""Base extractor for FoolFuuka based boards/archives""" """Base extractor for FoolFuuka based boards/archives"""
basecategory = "foolfuuka" basecategory = "foolfuuka"
filename_fmt = "{timestamp_ms} {filename_media}.{extension}"
archive_fmt = "{board[shortname]}_{num}_{timestamp}" archive_fmt = "{board[shortname]}_{num}_{timestamp}"
external = "default" external = "default"
@@ -40,6 +41,9 @@ class FoolfuukaExtractor(BaseExtractor):
post["filename"], _, post["extension"] = \ post["filename"], _, post["extension"] = \
media["media"].rpartition(".") media["media"].rpartition(".")
post["filename_media"] = media["media_filename"].rpartition(".")[0]
post["timestamp_ms"] = text.parse_int(
media["media_orig"].rpartition(".")[0])
yield Message.Url, url, post yield Message.Url, url, post
def metadata(self): def metadata(self):
@@ -107,7 +111,7 @@ class FoolfuukaThreadExtractor(FoolfuukaExtractor):
"""Base extractor for threads on FoolFuuka based boards/archives""" """Base extractor for threads on FoolFuuka based boards/archives"""
subcategory = "thread" subcategory = "thread"
directory_fmt = ("{category}", "{board[shortname]}", directory_fmt = ("{category}", "{board[shortname]}",
"{thread_num}{title:? - //}") "{thread_num} {title|comment[:50]}")
pattern = BASE_PATTERN + r"/([^/?#]+)/thread/(\d+)" pattern = BASE_PATTERN + r"/([^/?#]+)/thread/(\d+)"
test = ( test = (
("https://archive.4plebs.org/tg/thread/54059290", { ("https://archive.4plebs.org/tg/thread/54059290", {