X-Git-Url: https://git.decadent.org.uk/gitweb/?a=blobdiff_plain;f=dak%2Fcheck_archive.py;h=26a85cb5b42003ef68cfd84a5cd9b072b2ba302c;hb=4b277e613887f5103cd2ab64262fe29ad1bf7540;hp=7a2c779c8d4724dda07e9778d34f4936c5666cce;hpb=1d79ca7f4979f4b1d1b7741b4d2ab8af8db7c347;p=dak.git

diff --git a/dak/check_archive.py b/dak/check_archive.py
index 7a2c779c..26a85cb5 100755
--- a/dak/check_archive.py
+++ b/dak/check_archive.py
@@ -28,8 +28,9 @@
 
 import commands, os, pg, stat, sys, time
 import apt_pkg, apt_inst
-import daklib.database as database
-import daklib.utils as utils
+from daklib import database
+from daklib import utils
+from daklib.regexes import re_issource
 
 ################################################################################
 
@@ -52,7 +53,7 @@ Run various sanity checks of the archive and/or database.
 
 The following MODEs are available:
 
-  md5sums            - validate the md5sums stored in the database
+  checksums          - validate the checksums stored in the database
   files              - check files in the database against what's in the archive
   dsc-syntax         - validate the syntax of .dsc files in the archive
   missing-overrides  - check for missing overrides
@@ -194,16 +195,18 @@ SELECT l.path, f.filename FROM files f, dsc_files df, location l WHERE df.source
 
 ################################################################################
 
-def check_md5sums():
+def check_checksums():
     print "Getting file information from database..."
-    q = projectB.query("SELECT l.path, f.filename, f.md5sum, f.size FROM files f, location l WHERE f.location = l.id")
+    q = projectB.query("SELECT l.path, f.filename, f.md5sum, f.sha1sum, f.sha256sum, f.size FROM files f, location l WHERE f.location = l.id")
     ql = q.getresult()
 
-    print "Checking file md5sums & sizes..."
+    print "Checking file checksums & sizes..."
     for i in ql:
         filename = os.path.abspath(i[0] + i[1])
         db_md5sum = i[2]
-        db_size = int(i[3])
+        db_sha1sum = i[3]
+        db_sha256sum = i[4]
+        db_size = int(i[5])
         try:
             f = utils.open_file(filename)
         except:
@@ -215,6 +218,20 @@ def check_md5sums():
             utils.warn("**WARNING** md5sum mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, md5sum, db_md5sum))
         if size != db_size:
             utils.warn("**WARNING** size mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, size, db_size))
+        # Until the main database is filled, we need to not spit 500,000 warnings
+        # every time we scan the archive.  Yet another hack (TM) which can go away
+        # once this is all working
+        if db_sha1sum is not None and db_sha1sum != '':
+            f.seek(0)
+            sha1sum = apt_pkg.sha1sum(f)
+            if sha1sum != db_sha1sum:
+                utils.warn("**WARNING** sha1sum mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, sha1sum, db_sha1sum))
+
+        if db_sha256sum is not None and db_sha256sum != '':
+            f.seek(0)
+            sha256sum = apt_pkg.sha256sum(f)
+            if sha256sum != db_sha256sum:
+                utils.warn("**WARNING** sha256sum mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, sha256sum, db_sha256sum))
 
     print "Done."
 
@@ -271,7 +288,7 @@ def check_missing_tar_gz_in_dsc():
         dsc_files = utils.build_file_list(dsc, is_a_dsc=1)
         has_tar = 0
         for f in dsc_files.keys():
-            m = utils.re_issource.match(f)
+            m = re_issource.match(f)
             if not m:
                 utils.fubar("%s not recognised as source." % (f))
             ftype = m.group(3)
@@ -291,7 +308,7 @@ def validate_sources(suite, component):
     filename = "%s/dists/%s/%s/source/Sources.gz" % (Cnf["Dir::Root"], suite, component)
     print "Processing %s..." % (filename)
     # apt_pkg.ParseTagFile needs a real file handle and can't handle a GzipFile instance...
-    temp_filename = utils.temp_filename()
+    (fd, temp_filename) = utils.temp_filename()
     (result, output) = commands.getstatusoutput("gunzip -c %s > %s" % (filename, temp_filename))
     if (result != 0):
         sys.stderr.write("Gunzip invocation failed!\n%s\n" % (output))
@@ -330,7 +347,7 @@ def validate_packages(suite, component, architecture):
                % (Cnf["Dir::Root"], suite, component, architecture)
     print "Processing %s..." % (filename)
     # apt_pkg.ParseTagFile needs a real file handle and can't handle a GzipFile instance...
-    temp_filename = utils.temp_filename()
+    (fd, temp_filename) = utils.temp_filename()
     (result, output) = commands.getstatusoutput("gunzip -c %s > %s" % (filename, temp_filename))
     if (result != 0):
         sys.stderr.write("Gunzip invocation failed!\n%s\n" % (output))
@@ -425,8 +442,8 @@ def main ():
     projectB = pg.connect(Cnf["DB::Name"], Cnf["DB::Host"], int(Cnf["DB::Port"]))
     database.init(Cnf, projectB)
 
-    if mode == "md5sums":
-        check_md5sums()
+    if mode == "checksums":
+        check_checksums()
     elif mode == "files":
         check_files()
     elif mode == "dsc-syntax":