aboutsummaryrefslogtreecommitdiffstats
path: root/MgaRepo/binrepo.py
blob: 06575566821f189b423841a014bcc81eeb2dbac7 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
from MgaRepo import Error, config, mirror, layout
from MgaRepo.util import execcmd, rellink, get_helper
from MgaRepo.svn import SVN

import sys
import os
import string
import stat
import shutil
import re
import tempfile
import hashlib
import urlparse
import threading
import httplib2
from cStringIO import StringIO

DEFAULT_TARBALLS_REPO = "/tarballs"
BINARIES_DIR_NAME = "SOURCES"
BINARIES_CHECKOUT_NAME = "SOURCES-bin"

BINREPOS_SECTION = "binrepos"

SOURCES_FILE = "sha1.lst"

class ChecksumError(Error):
    pass

def svn_baseurl(target):
    svn = SVN()
    info = svn.info2(target)
    if info is None:
        # unversioned resource
        newtarget = os.path.dirname(target)
        info = svn.info2(newtarget)
        assert info is not None, "svn_basedir should not be used with a "\
                "non-versioned directory"
    root = info["Repository Root"]
    url = info["URL"]
    kind = info["Node Kind"]
    path = url[len(root):]
    if kind == "directory":
        return url
    basepath = os.path.dirname(path)
    baseurl = mirror.normalize_path(url + "/" + basepath)
    return baseurl

def default_repo():
    base = config.get("global", "binaries-repository", None)
    if base is None:
        default_parent = config.get("global", "default_parent", None)
        if default_parent is None:
            raise Error, "no binaries-repository nor default_parent "\
                    "configured"
        comps = urlparse.urlparse(default_parent)
        base = comps[1] + ":" + DEFAULT_TARBALLS_REPO
    return base

def translate_url(url):
    url = mirror.normalize_path(url)
    main = mirror.normalize_path(layout.repository_url())
    subpath = url[len(main)+1:]
    # [binrepos]
    # updates/2009.0 = svn+ssh://svn.mandriva.com/svn/binrepo/20090/
    ## svn+ssh://svn.mandriva.com/svn/packages/2009.0/trafshow/current
    ## would translate to 
    ## svn+ssh://svn.mandriva.com/svn/binrepo/20090/updates/trafshow/current/
    binbase = None
    if BINREPOS_SECTION in config.sections():
        for option, value in config.walk(BINREPOS_SECTION):
            if subpath.startswith(option):
                binbase = value
                break
    binurl = mirror._joinurl(binbase or default_repo(), subpath)
    return binurl

def translate_topdir(path):
    """Returns the URL in the binrepo from a given path inside a SVN
       checkout directory.

    @path: if specified, returns a URL in the binrepo whose path is the
           same as the path inside the main repository.
    """
    baseurl = svn_baseurl(path)
    binurl = translate_url(baseurl)
    target = mirror.normalize_path(binurl)
    return target

def is_binary(path):
    raw = config.get("binrepo", "upload-match",
            "\.(7z|Z|bin|bz2|cpio|db|deb|egg|gem|gz|jar|jisp|lzma|"\
               "pdf|pgn\\.gz|pk3|png|rpm|run|sdz|smzip|tar|tbz|"\
               "tbz2|tgz|ttf|uqm|wad|war|xar|xpi|xz|zip|wav|mp3|ogg|"\
	       "jpg|png|gif|avi|mpg|mpeg|rar)$")
    maxsize = config.getint("binrepo", "upload-match-size", "1048576") # 1MiB
    expr = re.compile(raw)
    name = os.path.basename(path)
    if expr.search(name):
        return True
    st = os.stat(path)
    if st[stat.ST_SIZE] >= maxsize:
        return True
    return False

def find_binaries(paths):
    new = []
    for path in paths:
        if os.path.isdir(path):
            for name in os.listdir(path):
                fpath = os.path.join(path, name)
                if is_binary(fpath):
                    new.append(fpath)
        else:
            if is_binary(path):
                new.append(path)
    return new

def download_binary(topdir, sha1, filename):
    fmt = config.get("global", "download-command",
	    "wget -c -O '$dest' $url")
    url = config.get("binrepo", "download_url",
	    "http://binrepo.mageia.org/")
    url = mirror.normalize_path(url + "/" + sha1)
    dest = os.path.join(topdir, 'SOURCES', filename)
    if os.path.exists(dest):
	return 1
    context = {"dest": dest, "url": url}
    try:
	cmd = string.Template(fmt).substitute(context)
    except KeyError, e:
	raise Error, "invalid variable %r in download-command "\
		"configuration option" % e
    try:
	status, output = execcmd(cmd, show=True)
    except Error, e:
	raise Error, "Could not download file %s\n" % url

def download_binaries(topdir):
    spath = sources_path(topdir)
    if not os.path.exists(spath):
        raise Error, "'%s' was not found" % spath
    entries = parse_sources(spath)
    for name, sha1 in entries.iteritems():
	download_binary(topdir, sha1, name)

def binary_exists(sha1sum):
    dlurl = config.get("binrepo", "download_url",
	    "http://binrepo.mageia.org/")
    dlurl = mirror.normalize_path(dlurl + "/" + sha1sum)
    h = httplib2.Http()
    resp, content = h.request(dlurl, 'HEAD')
    return resp.status == 200

def upload_binary(topdir, filename):
    filepath = os.path.join(topdir, 'SOURCES', filename)
    if not os.path.exists(filepath):
        raise Error, "'%s' was not found" % spath
    sha1sum = file_hash(filepath)
    if binary_exists(sha1sum):
	return
    host = config.get("binrepo", "upload_host")
    upload_bin_helper = get_helper("upload-bin")
    command = "ssh %s %s %s" % (host, upload_bin_helper, filename)
    try:
	filein = open(filepath, 'r')
    except Error, e:
	raise Error, "Could not open file %s\n" % filepath
    status, output = execcmd(command, show=True, geterr=True, stdin=filein)

def import_binaries(topdir, pkgname):
    """Import all binaries from a given package checkout

    (with pending svn adds)

    @topdir: the path to the svn checkout
    """
    svn = SVN()
    topurl = translate_topdir(topdir)
    sourcesdir = os.path.join(topdir, "SOURCES")
    bintopdir = tempfile.mktemp("repsys")
    try:
        svn.checkout(topurl, bintopdir)
        checkout = True
    except Error:
        bintopdir = tempfile.mkdtemp("repsys")
        checkout = False
    try:
        bindir = os.path.join(bintopdir, BINARIES_DIR_NAME)
        if not os.path.exists(bindir):
            if checkout:
                svn.mkdir(bindir)
            else:
                os.mkdir(bindir)
        binaries = find_binaries([sourcesdir])
        update = update_sources_threaded(topdir, added=binaries)
        for path in binaries:
            name = os.path.basename(path)
            binpath = os.path.join(bindir, name)
            os.rename(path, binpath)
            try:
                svn.remove(path)
            except Error:
                # file not tracked
                svn.revert(path)
            if checkout:
                svn.add(binpath)
        log = "imported binaries for %s" % pkgname
        if checkout:
            rev = svn.commit(bindir, log=log)
        else:
            rev = svn.import_(bintopdir, topurl, log=log)
        update.join()
        svn.add(sources_path(topdir))
    finally:
        shutil.rmtree(bintopdir)

def parse_sources(path):
    entries = {}
    f = open(path)
    for rawline in f:
        line = rawline.strip()
        try:
            sum, name = line.split(None, 1)
        except ValueError:
            # failed to unpack, line format error
            raise Error, "invalid line in sources file: %s" % rawline
        entries[name] = sum
    return entries

def check_hash(path, sum):
    newsum = file_hash(path)
    if newsum != sum:
        raise ChecksumError, "different checksums for %s: expected %s, "\
                "but %s was found" % (path, sum, newsum)

def check_sources(topdir):
    spath = sources_path(topdir)
    if not os.path.exists(spath):
        raise Error, "'%s' was not found" % spath
    entries = parse_sources(spath)
    for name, sum in entries.iteritems():
        fpath = os.path.join(topdir, "SOURCES", name)
        check_hash(fpath, sum)

def file_hash(path):
    sum = hashlib.sha1()
    f = open(path)
    while True:
        block = f.read(4096)
        if not block:
            break
        sum.update(block)
    f.close()
    return sum.hexdigest()

def sources_path(topdir):
    path = os.path.join(topdir, "SOURCES", SOURCES_FILE)
    return path

def update_sources(topdir, added=[], removed=[]):
    path = sources_path(topdir)
    entries = {}
    if os.path.isfile(path):
        entries = parse_sources(path)
    f = open(path, "w") # open before calculating hashes
    for name in removed:
	if name in entries:
           del entries[name]
    for added_path in added:
        name = os.path.basename(added_path)
        entries[name] = file_hash(added_path)
    for name in sorted(entries):
        f.write("%s  %s\n" % (entries[name], name))
    f.close()

def update_sources_threaded(*args, **kwargs):
    t = threading.Thread(target=update_sources, args=args, kwargs=kwargs)
    t.start()
    t.join()
    return t

def mapped_revision(target, revision, wc=False):
    """Maps a txtrepo revision to a binrepo datespec

    This datespec can is intended to be used by svn .. -r DATE.

    @target: a working copy path or a URL
    @revision: if target is a URL, the revision number used when fetching
         svn info
    @wc: if True indicates that 'target' must be interpreted as a
         the path of a svn working copy, otherwise it is handled as a URL
    """
    svn = SVN()
    binrev = None
    if wc:
        spath = sources_path(target)
        if os.path.exists(spath):
            infolines = svn.info(spath, xml=True)
            if infolines:
                rawinfo = "".join(infolines) # arg!
                found = re.search("<date>(.*?)</date>", rawinfo).groups()
                date = found[0]
            else:
                raise Error, "bogus 'svn info' for '%s'" % spath
        else:
            raise Error, "'%s' was not found" % spath
    else:
        url = mirror._joinurl(target, sources_path(""))
        date = svn.propget("svn:date", url, rev=revision, revprop=True)
        if not date:
            raise Error, "no valid date available for '%s'" % url
    binrev = "{%s}" % date
    return binrev

def markrelease(sourceurl, releasesurl, version, release, revision, binrev):
    svn = SVN()
    if not binrev:
	binrev = mapped_revision(sourceurl, revision)
    binsource = translate_url(sourceurl)
    binreleases = translate_url(releasesurl)
    versiondir = mirror._joinurl(binreleases, version)
    dest = mirror._joinurl(versiondir, release)
    svn.mkdir(binreleases, noerror=1, log="created directory for releases")
    svn.mkdir(versiondir, noerror=1, log="created directory for version %s" % version)
    svn.copy(binsource, dest, rev=binrev,
            log="%%markrelease ver=%s rel=%s rev=%s binrev=%s" % (version, release,
                revision, binrev))