diff options
author | Sebastian Thiel <byronimo@gmail.com> | 2009-10-11 22:50:44 +0200 |
---|---|---|
committer | Sebastian Thiel <byronimo@gmail.com> | 2009-10-11 22:50:44 +0200 |
commit | 3c0a65226f038c58fc6d6ed525f38fc00b3579b7 (patch) | |
tree | a5b715a490d9cbd8f45eabc1968374c96bdea1c0 /lib/git/blob.py | |
parent | 9c0c2fc4ee2d8a5d0a2de50ba882657989dedc51 (diff) | |
parent | c68459a17ff59043d29c90020fffe651b2164e6a (diff) | |
download | gitpython-3c0a65226f038c58fc6d6ed525f38fc00b3579b7.tar.gz |
Merge branch 'hierarchyfix' into improvements
* hierarchyfix:
Added remaining tests for new base classes and removed some methods whose existance was doubtful or unsafe
Fixed remaining tests to deal with the changes
commit: fixed failing commit tests as the mocked git command would always return the same thing which does not work anymore - re-implemented it in a more dynamic manner, but in the end tests will have to be revised anyway
mode-only change for test system - this should be in a separate repository in fact so that changes are a little more self-contained and not depending on the actual source repository
fixed issue in Ref.name implementation which would not handle components properly
lazymixin system now supports per-attribute baking, it is up to the class whether it bakes more. This also leads to more efficient use of memory as values are only cached and set when required - the baking system does not require an own tracking variable anymore, and values are only to be cached once - then python will natively find the cache without involving any additional overhead. This works by using __getattr__ instead of __get_attribute__ which would always be called
put Tree and Blob onto a new base class suitable to deal with IndexObjects
blob tests fixed to deal with changes to the Blob type
converted all spaces to tabs ( 4 spaces = 1 tab ) just to allow me and my editor to work with the files properly. Can convert it back for releaes
Re-designed the tag testing - it does not use fixtures anymore but dyamically checks the existance of tags within the repository - it basically tests the interface and checks that expected return types are actually returned
Intermediate commit: commit,tree and blob objects now derive from object - test is in place which still fails on purpose. Need to integrate tags which can be objects or just a special form of a ref
Renamed lazy.py to base.py to have a file for base classes - lazy not yet changed to allow proper rename tracking
Diffstat (limited to 'lib/git/blob.py')
-rw-r--r-- | lib/git/blob.py | 282 |
1 files changed, 111 insertions, 171 deletions
diff --git a/lib/git/blob.py b/lib/git/blob.py index 1e8aa12b..1fafb128 100644 --- a/lib/git/blob.py +++ b/lib/git/blob.py @@ -10,177 +10,117 @@ import re import time from actor import Actor from commit import Commit - -class Blob(object): - """A Blob encapsulates a git blob object""" - DEFAULT_MIME_TYPE = "text/plain" - - # precompiled regex - re_whitespace = re.compile(r'\s+') - re_hexsha_only = re.compile('^[0-9A-Fa-f]{40}$') - re_author_committer_start = re.compile(r'^(author|committer)') - re_tab_full_line = re.compile(r'^\t(.*)$') - - def __init__(self, repo, id, mode=None, path=None): - """ - Create an unbaked Blob containing just the specified attributes - - ``repo`` - is the Repo - - ``id`` - is the git object id - - ``mode`` - is the file mode - - ``path`` - is the path to the file - - Returns - git.Blob - """ - self.repo = repo - self.id = id - self.mode = mode - self.path = path - - self._size = None - self.data_stored = None - - @property - def size(self): - """ - The size of this blob in bytes - - Returns - int - - NOTE - The size will be cached after the first access - """ - if self._size is None: - self._size = int(self.repo.git.cat_file(self.id, s=True).rstrip()) - return self._size - - @property - def data(self): - """ - The binary contents of this blob. - - Returns - str - - NOTE - The data will be cached after the first access. - """ - self.data_stored = self.data_stored or self.repo.git.cat_file(self.id, p=True, with_raw_output=True) - return self.data_stored - - @property - def mime_type(self): - """ - The mime type of this file (based on the filename) - - Returns - str - - NOTE - Defaults to 'text/plain' in case the actual file type is unknown. - """ - guesses = None - if self.path: - guesses = mimetypes.guess_type(self.path) - return guesses and guesses[0] or self.DEFAULT_MIME_TYPE - - @property - def basename(self): - """ - Returns - The basename of the Blobs file path - """ - return os.path.basename(self.path) - - @classmethod - def blame(cls, repo, commit, file): - """ - The blame information for the given file at the given commit - - Returns - list: [git.Commit, list: [<line>]] - A list of tuples associating a Commit object with a list of lines that - changed within the given commit. The Commit objects will be given in order - of appearance. - """ - data = repo.git.blame(commit, '--', file, p=True) - commits = {} - blames = [] - info = None - - for line in data.splitlines(): - parts = cls.re_whitespace.split(line, 1) - firstpart = parts[0] - if cls.re_hexsha_only.search(firstpart): - # handles - # 634396b2f541a9f2d58b00be1a07f0c358b999b3 1 1 7 - indicates blame-data start - # 634396b2f541a9f2d58b00be1a07f0c358b999b3 2 2 - digits = parts[-1].split(" ") - if len(digits) == 3: +import base + +class Blob(base.IndexObject): + """A Blob encapsulates a git blob object""" + DEFAULT_MIME_TYPE = "text/plain" + type = "blob" + + __slots__ = tuple() + + # precompiled regex + re_whitespace = re.compile(r'\s+') + re_hexsha_only = re.compile('^[0-9A-Fa-f]{40}$') + re_author_committer_start = re.compile(r'^(author|committer)') + re_tab_full_line = re.compile(r'^\t(.*)$') + + @property + def mime_type(self): + """ + The mime type of this file (based on the filename) + + Returns + str + + NOTE + Defaults to 'text/plain' in case the actual file type is unknown. + """ + guesses = None + if self.path: + guesses = mimetypes.guess_type(self.path) + return guesses and guesses[0] or self.DEFAULT_MIME_TYPE + + @classmethod + def blame(cls, repo, commit, file): + """ + The blame information for the given file at the given commit + + Returns + list: [git.Commit, list: [<line>]] + A list of tuples associating a Commit object with a list of lines that + changed within the given commit. The Commit objects will be given in order + of appearance. + """ + data = repo.git.blame(commit, '--', file, p=True) + commits = {} + blames = [] + info = None + + for line in data.splitlines(False): + parts = cls.re_whitespace.split(line, 1) + firstpart = parts[0] + if cls.re_hexsha_only.search(firstpart): + # handles + # 634396b2f541a9f2d58b00be1a07f0c358b999b3 1 1 7 - indicates blame-data start + # 634396b2f541a9f2d58b00be1a07f0c358b999b3 2 2 + digits = parts[-1].split(" ") + if len(digits) == 3: info = {'id': firstpart} blames.append([None, []]) # END blame data initialization - else: - m = cls.re_author_committer_start.search(firstpart) - if m: - # handles: - # author Tom Preston-Werner - # author-mail <tom@mojombo.com> - # author-time 1192271832 - # author-tz -0700 - # committer Tom Preston-Werner - # committer-mail <tom@mojombo.com> - # committer-time 1192271832 - # committer-tz -0700 - IGNORED BY US - role = m.group(0) - if firstpart.endswith('-mail'): - info["%s_email" % role] = parts[-1] - elif firstpart.endswith('-time'): - info["%s_date" % role] = time.gmtime(int(parts[-1])) - elif role == firstpart: - info[role] = parts[-1] - # END distinguish mail,time,name - else: - # handle - # filename lib/grit.rb - # summary add Blob - # <and rest> - if firstpart.startswith('filename'): - info['filename'] = parts[-1] - elif firstpart.startswith('summary'): - info['summary'] = parts[-1] - elif firstpart == '': - if info: - sha = info['id'] - c = commits.get(sha) - if c is None: - c = Commit( repo, id=sha, - author=Actor.from_string(info['author'] + ' ' + info['author_email']), - authored_date=info['author_date'], - committer=Actor.from_string(info['committer'] + ' ' + info['committer_email']), - committed_date=info['committer_date'], - message=info['summary']) - commits[sha] = c - # END if commit objects needs initial creation - m = cls.re_tab_full_line.search(line) - text, = m.groups() - blames[-1][0] = c - blames[-1][1].append( text ) - info = None - # END if we collected commit info - # END distinguish filename,summary,rest - # END distinguish author|committer vs filename,summary,rest - # END distinguish hexsha vs other information - return blames - - def __repr__(self): - return '<git.Blob "%s">' % self.id + else: + m = cls.re_author_committer_start.search(firstpart) + if m: + # handles: + # author Tom Preston-Werner + # author-mail <tom@mojombo.com> + # author-time 1192271832 + # author-tz -0700 + # committer Tom Preston-Werner + # committer-mail <tom@mojombo.com> + # committer-time 1192271832 + # committer-tz -0700 - IGNORED BY US + role = m.group(0) + if firstpart.endswith('-mail'): + info["%s_email" % role] = parts[-1] + elif firstpart.endswith('-time'): + info["%s_date" % role] = time.gmtime(int(parts[-1])) + elif role == firstpart: + info[role] = parts[-1] + # END distinguish mail,time,name + else: + # handle + # filename lib/grit.rb + # summary add Blob + # <and rest> + if firstpart.startswith('filename'): + info['filename'] = parts[-1] + elif firstpart.startswith('summary'): + info['summary'] = parts[-1] + elif firstpart == '': + if info: + sha = info['id'] + c = commits.get(sha) + if c is None: + c = Commit( repo, id=sha, + author=Actor.from_string(info['author'] + ' ' + info['author_email']), + authored_date=info['author_date'], + committer=Actor.from_string(info['committer'] + ' ' + info['committer_email']), + committed_date=info['committer_date'], + message=info['summary']) + commits[sha] = c + # END if commit objects needs initial creation + m = cls.re_tab_full_line.search(line) + text, = m.groups() + blames[-1][0] = c + blames[-1][1].append( text ) + info = None + # END if we collected commit info + # END distinguish filename,summary,rest + # END distinguish author|committer vs filename,summary,rest + # END distinguish hexsha vs other information + return blames + + def __repr__(self): + return '<git.Blob "%s">' % self.id |