paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-09-16 21:55:37 -05:00
Code Issues Packages Projects Releases Wiki Activity

Add unitest for tika_parse()

Browse Source
This commit is contained in:
phail
2022-10-15 13:13:29 +02:00
parent 3d37e49c1a
commit daf90399bd
3 changed files with 226 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
src/paperless_mail/parsers.py
View File

@@ -21,6 +21,7 @@ class MailDocumentParser(DocumentParser):
"""
gotenberg_server = settings.PAPERLESS_TIKA_GOTENBERG_ENDPOINT
tika_server = settings.PAPERLESS_TIKA_ENDPOINT
logging_name = "paperless.parsing.mail"
_parsed = None
@@ -133,13 +134,13 @@ class MailDocumentParser(DocumentParser):
def tika_parse(self, html: str):
self.log("info", "Sending content to Tika server")
tika_server = settings.PAPERLESS_TIKA_ENDPOINT
try:
parsed = parser.from_buffer(html, tika_server)
parsed = parser.from_buffer(html, self.tika_server)
except Exception as err:
raise ParseError(
f"Could not parse content with tika server at " f"{tika_server}: {err}",
f"Could not parse content with tika server at "
f"{self.tika_server}: {err}",
)
if parsed["content"]:
return parsed["content"]
@@ -246,7 +247,7 @@ class MailDocumentParser(DocumentParser):
html = StringIO()
with open(html_file, "r") as html_template_handle:
with open(html_file) as html_template_handle:
for line in html_template_handle.readlines():
for placeholder in placeholder_pattern.findall(line):
line = re.sub(