Use a collections.defaultdict to avoid boilerplate definitions.

[dak.git] / daklib / utils.py

diff --git a/daklib/utils.py b/daklib/utils.py
index 5edf2cb1eeddcd34a3dd7f5237b029a3f863ba8c..c4f93fe5f651cc2b8bc5c6b82861a548f554aebc 100755 (executable)
--- a/daklib/utils.py
+++ b/daklib/utils.py
@@ -45,8 +45,11 @@ from dak_exceptions import *
 from textutils import fix_maintainer
 from regexes import re_html_escaping, html_escaping, re_single_line_field, \
                     re_multi_line_field, re_srchasver, re_verwithext, \
-                    re_parse_maintainer, re_taint_free, re_gpg_uid, re_re_mark, \
-                    re_whitespace_comment
+                    re_parse_maintainer, re_taint_free, re_gpg_uid, \
+                    re_re_mark, re_whitespace_comment, re_issource
+
+from srcformats import srcformats
+from collections import defaultdict
 
 ################################################################################
 
@@ -332,6 +335,79 @@ def check_size(where, files):
 
 ################################################################################
 
+def check_dsc_files(dsc_filename, dsc=None, dsc_files=None):
+    """
+    Verify that the files listed in the Files field of the .dsc are
+    those expected given the announced Format.
+
+    @type dsc_filename: string
+    @param dsc_filename: path of .dsc file
+
+    @type dsc: dict
+    @param dsc: the content of the .dsc parsed by C{parse_changes()}
+
+    @type dsc_files: dict
+    @param dsc_files: the file list returned by C{build_file_list()}
+
+    @rtype: list
+    @return: all errors detected
+    """
+    rejmsg = []
+
+    # Parse the file if needed
+    if dsc is None:
+        dsc = parse_changes(dsc_filename, signing_rules=1);
+
+    if dsc_files is None:
+        dsc_files = build_file_list(dsc, is_a_dsc=1)
+
+    # Ensure .dsc lists proper set of source files according to the format
+    # announced
+    has = defaultdict(lambda: 0)
+
+    for f in dsc_files.keys():
+        m = re_issource.match(f)
+        if not m:
+            rejmsg.append("%s: %s in Files field not recognised as source."
+                          % (dsc_filename, f))
+            continue
+        ftype = m.group(3)
+        if ftype == "orig.tar.gz":
+            has['orig_tar_gz'] += 1
+            has['orig_tar'] += 1
+        elif ftype == "diff.gz":
+            has['debian_diff'] += 1
+        elif ftype == "tar.gz":
+            has['native_tar_gz'] += 1
+            has['native_tar'] += 1
+        elif re.match(r"debian\.tar\.(gz|bz2|lzma)", ftype):
+            has['debian_tar'] += 1
+        elif re.match(r"orig\.tar\.(gz|bz2|lzma)", ftype):
+            has['orig_tar'] += 1
+        elif re.match(r"tar\.(gz|bz2|lzma)", ftype):
+            has['native_tar'] += 1
+        elif re.match(r"orig-.+\.tar\.(gz|bz2|lzma)", ftype):
+            has['more_orig_tar'] += 1
+        else:
+            reject("%s: unexpected source file '%s'" % (dsc_filename, f))
+    if has['orig_tar'] > 1:
+        rejmsg.append("%s: lists multiple .orig tarballs." % (dsc_filename))
+    if has['native_tar'] > 1:
+        rejmsg.append("%s: lists multiple native tarballs." % (dsc_filename))
+    if has['debian_tar'] > 1 or has['debian_diff'] > 1:
+        rejmsg.append("%s: lists multiple debian diff/tarballs." % (dsc_filename))
+
+    for format in srcformats:
+        if format.re_format.match(dsc['format']):
+            rejmsg.extend([
+                '%s: %s' % (dsc_filename, x) for x in format.reject_msgs(has)
+            ])
+            break
+
+    return rejmsg
+
+################################################################################
+
 def check_hash_fields(what, manifest):
     """
     check_hash_fields ensures that there are no checksum fields in the
@@ -394,7 +470,12 @@ def parse_checksums(where, files, manifest, hashname):
     for line in manifest[field].split('\n'):
         if not line:
             break
-        checksum, size, checkfile = line.strip().split(' ')
+        clist = line.strip().split(' ')
+        if len(clist) == 3:
+            checksum, size, checkfile = clist
+        else:
+            rejmsg.append("Cannot parse checksum line [%s]" % (line))
+            continue
         if not files.has_key(checkfile):
         # TODO: check for the file's entry in the original files dict, not
         # the one modified by (auto)byhand and other weird stuff
@@ -437,10 +518,10 @@ def build_file_list(changes, is_a_dsc=0, field="files", hashname="md5sum"):
         format = format[:2]
 
     if is_a_dsc:
-        # format = (1,0) are the only formats we currently accept,
         # format = (0,0) are missing format headers of which we still
         # have some in the archive.
-        if format != (1,0) and format != (0,0):
+        if format != (1,0) and format != (0,0) and \
+           format != (3,0,"quilt") and format != (3,0,"native"):
             raise UnknownFormatError, "%s" % (changes.get("format","0.0"))
     else:
         if (format < (1,5) or format > (1,8)):
@@ -671,7 +752,7 @@ def TemplateSubst(map, filename):
     templatefile = open_file(filename)
     template = templatefile.read()
     for x in map.keys():
-        template = template.replace(x,map[x])
+        template = template.replace(x, str(map[x]))
     templatefile.close()
     return template