regression-hotspots: allow tdf1234 style IDs, exclude some components

Change-Id: Id7ab2bd48a3fa1047798d28e102c1d8059eadd17
Reviewed-on: https://gerrit.libreoffice.org/c/dev-tools/+/166030
Tested-by: Ilmari Lauhakangas <ilmari.lauhakangas@libreoffice.org>
Reviewed-by: Ilmari Lauhakangas <ilmari.lauhakangas@libreoffice.org>

1 files changed, 4 insertions, 3 deletions

diff --git a/scripts/regression-hotspots.py b/scripts/regression-hotspots.py
index 27a684fe..150cb9bd 100755
--- a/scripts/regression-hotspots.py
+++ b/scripts/regression-hotspots.py
@@ -19,7 +19,7 @@ from urllib.request import urlopen, URLError
 from io import BytesIO
 
 def get_fixed_regression_bugs():
-    url = 'https://bugs.documentfoundation.org/buglist.cgi?columnlist=&keywords=regression%2C%20&keywords_type=allwords&limit=0&product=LibreOffice&resolution=FIXED&ctype=csv&human=0'
+    url = 'https://bugs.documentfoundation.org/buglist.cgi?f1=component&f2=component&f3=component&f4=component&f5=component&n1=1&n2=1&n3=1&n4=1&n5=1&o1=equals&o2=equals&o3=equals&o4=equals&o5=equals&v1=ci-infra&v2=deletionRequest&v3=FirefoxOS%20app&v4=SI-GUI&v5=WWW&columnlist=&keywords=regression%2C%20&keywords_type=allwords&limit=0&product=LibreOffice&resolution=FIXED&ctype=csv&human=0'
 
     ctx = ssl.create_default_context()
     ctx.check_hostname = False
@@ -63,14 +63,15 @@ if __name__ == '__main__':
 
     # build a dictionary of hashes and bug IDs from all commits targeting a report in FDO/TDF Bugzilla
     # (first commit with fdo# aka freedesktop.org is from 1 Oct 2010)
+    # sometimes people accidentally leave out the #, so take that into account in the regexes
     gitbugs = {}
-    buglog = git.Git('.').execute(['git', 'log', '--grep=(fdo|tdf)#', '-E', '--oneline', '--since=1.10.2010'])
+    buglog = git.Git('.').execute(['git', 'log', '--grep=(fdo|tdf)#*', '-E', '--oneline', '--since=1.10.2010'])
     if buglog:
         for line in buglog.split('\n'):
             githash = line.partition(' ')[0]
             # the regex search will ignore any commits hit by the grep where fdo|tdf# occurred below
             # the first line - this is desirable as the referred bug ID should appear in the subject line
-            bugid = re.search(r"(?:fdo|tdf)#([0-9]+)", line)
+            bugid = re.search(r"(?:fdo|tdf)#*([0-9]+)", line)
             if bugid:
                 gitbugs[githash] = int(bugid.group(1))