viewcvs.py

PHP写的图片查看器

# -*-python-*-
#
# Copyright (C) 1999-2001 The ViewCVS Group. All Rights Reserved.
#
# By using this file, you agree to the terms and conditions set forth in
# the LICENSE.html file which can be found at the top level of the ViewCVS
# distribution or at http://viewcvs.sourceforge.net/license-1.html.
#
# Contact information:
#   Greg Stein, PO Box 760, Palo Alto, CA, 94302
#   [e-mail], http://viewcvs.sourceforge.net/
#
# -----------------------------------------------------------------------
#
# viewcvs: View CVS repositories via a web browser
#
# -----------------------------------------------------------------------
#
# This software is based on "cvsweb" by Henner Zeller (which is, in turn,
# derived from software by Bill Fenner, with additional modifications by
# Henrik Nordstrom and Ken Coar). The cvsweb distribution can be found
# on Zeller's site:
#   http://stud.fh-heilbronn.de/~zeller/cgi/cvsweb.cgi/
#
# -----------------------------------------------------------------------
#

__version__ = '0.9.2'

#########################################################################
#
# INSTALL-TIME CONFIGURATION
#
# These values will be set during the installation process. During
# development, they will remain None.
#

CONF_PATHNAME = None

#########################################################################

# this comes from our library; measure the startup time
import debug
debug.t_start('startup')
debug.t_start('imports')

# standard modules that we know are in the path or builtin
import sys
import os
import cgi
import string
import urllib
import mimetypes
import time
import re
import stat
import struct

# these modules come from our library (the stub has set up the path)
import compat
import config
import popen
import ezt
import accept

debug.t_end('imports')

#########################################################################

checkout_magic_path = '*checkout*'
# According to RFC 1738 the '~' character is unsafe in URLs.
# But for compatibility with URLs bookmarked with older releases of ViewCVS:
oldstyle_checkout_magic_path = '~checkout~'
docroot_magic_path = '*docroot*'
viewcvs_mime_type = 'text/vnd.viewcvs-markup'

# put here the variables we need in order to hold our state - they will be
# added (with their current value) to any link/query string you construct
_sticky_vars = (
  'cvsroot',
  'hideattic',
  'sortby',
  'logsort',
  'diff_format',
  'only_with_tag',
  'search',
  )

# regex used to move from a file to a directory
_re_up_path = re.compile('(Attic/)?[^/]+$')

_EOF_FILE = 'end of file entries'       # no more entries for this RCS file
_EOF_LOG = 'end of log'                 # hit the true EOF on the pipe
_EOF_ERROR = 'error message found'      # rlog issued an error

_FILE_HAD_ERROR = 'could not read file'

_UNREADABLE_MARKER = '//UNREADABLE-MARKER//'

# for reading/writing between a couple descriptors
CHUNK_SIZE = 8192

# if your rlog doesn't use 77 '=' characters, then this must change
LOG_END_MARKER = '=' * 77 + '\n'
ENTRY_END_MARKER = '-' * 28 + '\n'

# global configuration:
cfg = None # see below

if CONF_PATHNAME:
  # installed
  g_install_dir = os.path.dirname(CONF_PATHNAME)
else:
  # development directories
  g_install_dir = os.pardir # typically, ".."


class Request:
  def __init__(self):
    where = os.environ.get('PATH_INFO', '')

    # clean it up. this removes duplicate '/' characters and any that may
    # exist at the front or end of the path.
    parts = filter(None, string.split(where, '/'))

    self.has_checkout_magic = 0
    self.has_docroot_magic = 0
    # does it have a magic prefix?
    if parts:
      if parts[0] in (checkout_magic_path, oldstyle_checkout_magic_path):
        self.has_checkout_magic = 1
        del parts[0]
      elif parts[0] == docroot_magic_path:
        self.has_docroot_magic = 1
        del parts[0]

    # put it back together
    where = string.join(parts, '/')

    script_name = os.environ['SCRIPT_NAME']     ### clean this up?
    if where:
      url = script_name + '/' + urllib.quote(where)
    else:
      url = script_name

    self.where = where
    self.script_name = script_name
    self.url = url
    if parts:
      self.module = parts[0]
    else:
      self.module = None

    self.browser = os.environ.get('HTTP_USER_AGENT', 'unknown')

    # in lynx, it it very annoying to have two links
    # per file, so disable the link at the icon
    # in this case:
    self.no_file_links = string.find(self.browser, 'Lynx') != -1

    # newer browsers accept gzip content encoding
    # and state this in a header
    # (netscape did always but didn't state it)
    # It has been reported that these
    #  braindamaged MS-Internet Explorers claim that they
    # accept gzip .. but don't in fact and
    # display garbage then :-/
    self.may_compress = (
      ( string.find(os.environ.get('HTTP_ACCEPT_ENCODING', ''), 'gzip') != -1
        or string.find(self.browser, 'Mozilla/3') != -1)
      and string.find(self.browser, 'MSIE') == -1
      )

    # parse the query params into a dictionary (and use defaults)
    query_dict = default_settings.copy()
    for name, values in cgi.parse().items():
      query_dict[name] = values[0]

    # set up query strings, prefixed by question marks and ampersands
    query = sticky_query(query_dict)
    if query:
      self.qmark_query = '?' + query
      self.amp_query = '&' + query
    else:
      self.qmark_query = ''
      self.amp_query = ''

    self.query_dict = query_dict

    # set up the CVS repository to use
    self.cvsrep = query_dict.get('cvsroot', cfg.general.default_root)
    try:
      self.cvsroot = cfg.general.cvs_roots[self.cvsrep]
    except KeyError:
      if query_dict.has_key('cvsroot'):
          error("Repository cvsroot %s not configured in viewcvs.conf" % 
                self.cvsrep, "404 Repository not found")
      else:
          error("The settings of 'cvs_roots' and "
                "default_root=%s are misconfigured in the viewcvs.conf file." %
                self.cvsrep)

    self.full_name = self.cvsroot + '/' + where

    # process the Accept-Language: header
    hal = os.environ.get('HTTP_ACCEPT_LANGUAGE')
    self.lang_selector = accept.language(hal)
    self.language = self.lang_selector.select_from(cfg.general.languages)

    # load the key/value files, given the selected language
    self.kv = cfg.load_kv_files(self.language)

  def setup_mime_type_info(self):
    if cfg.general.mime_types_file:
      mimetypes.init([cfg.general.mime_types_file])
    self.mime_type, self.encoding = mimetypes.guess_type(self.where)
    if not self.mime_type:
      self.mime_type = 'text/plain'
    self.default_viewable = cfg.options.allow_markup and \
                            is_viewable(self.mime_type)


class LogHeader:
  "Hold state from the header portion of an 'rlog' output."
  def __init__(self, filename, head=None, branch=None, taginfo=None):
    self.filename = filename
    self.head = head
    self.branch = branch
    self.taginfo = taginfo


class LogEntry:
  "Hold state for each revision entry in an 'rlog' output."
  def __init__(self, rev, date, author, state, changed, log):
    self.rev = rev
    self.date = date
    self.author = author
    self.state = state
    self.changed = changed
    self.log = log


def redirect(location):
  print 'Status: 301 Moved'
  print 'Location:', location
  print
  print 'This document is located <a href="%s">here</a>.' % location
  sys.exit(0)

def error(msg, status='500 Internal Server Error'):
  print 'Status:', status
  print
  print msg
  sys.exit(0)

def generate_page(request, tname, data):
  # allow per-language template selection
  if request:
    tname = string.replace(tname, '%lang%', request.language)
  else:
    tname = string.replace(tname, '%lang%', 'en')

  debug.t_start('ezt-parse')
  template = ezt.Template(os.path.join(g_install_dir, tname))
  debug.t_end('ezt-parse')

  template.generate(sys.stdout, data)

_header_sent = 0
def http_header(content_type='text/html'):
  global _header_sent
  if _header_sent:
    return
  print 'Content-Type:', content_type
  print
  _header_sent = 1

def html_footer(request):
  ### would be nice to have a "standard" set of data available to all
  ### templates. should move that to the request ob, probably
  data = {
    'cfg' : cfg,
    'vsn' : __version__,
    }

  if request:
    data['kv'] = request.kv

  # generate the footer
  generate_page(request, cfg.templates.footer, data)

def sticky_query(dict):
  sticky_dict = { }
  for varname in _sticky_vars:
    value = dict.get(varname)
    if value is not None and value != default_settings.get(varname, ''):
      sticky_dict[varname] = value
  return compat.urlencode(sticky_dict)

def toggle_query(query_dict, which, value=None):
  dict = query_dict.copy()
  if value is None:
    dict[which] = not dict[which]
  else:
    dict[which] = value
  query = sticky_query(dict)
  if query:
    return '?' + query
  return ''

def clickable_path(request, path, leaf_is_link, leaf_is_file, drop_leaf):
  s = '<a href="%s/%s#dirlist">[%s]</a>' % \
      (request.script_name, request.qmark_query, request.cvsrep)
  parts = filter(None, string.split(path, '/'))
  if drop_leaf:
    del parts[-1]
    leaf_is_link = 1
    leaf_is_file = 0
  where = ''
  for i in range(len(parts)):
    where = where + '/' + parts[i]
    is_leaf = i == len(parts) - 1
    if not is_leaf or leaf_is_link:
      if is_leaf and leaf_is_file:
        slash = ''
      else:
        slash = '/'
      ### should we be encoding/quoting the URL stuff? (probably...)
      s = s + ' / <a href="%s%s%s%s#dirlist">%s</a>' % \
          (request.script_name, where, slash, request.qmark_query, parts[i])
    else:
      s = s + ' / ' + parts[i]

  return s

def prep_tags(query_dict, file_url, tags):
  links = [ ]
  for tag in tags:
    href = file_url + toggle_query(query_dict, 'only_with_tag', tag)
    links.append(_item(name=tag, href=href))
  return links

def is_viewable(mime_type):
  return mime_type[:5] == 'text/' or mime_type[:6] == 'image/'

_re_rewrite_url = re.compile('((http|ftp)(://[-a-zA-Z0-9%.~:_/]+)([?&]([-a-zA-Z0-9%.~:_]+)=([-a-zA-Z0-9%.~:_])+)*(#([-a-zA-Z0-9%.~:_]+)?)?)')
_re_rewrite_email = re.compile('([-a-zA-Z0-9_.]+@([-a-zA-Z0-9]+\.)+[A-Za-z]{2,4})')
def htmlify(html):
  html = cgi.escape(html)
  html = re.sub(_re_rewrite_url, r'<a href="\1">\1</a>', html)
  html = re.sub(_re_rewrite_email, r'<a href="mailto:\1">\1</a>', html)
  return html

def format_log(log):
  s = htmlify(log[:cfg.options.short_log_len])
  if len(log) > cfg.options.short_log_len:
    s = s + '...'
  return s

def download_url(request, url, revision, mime_type):
  if cfg.options.checkout_magic and mime_type != viewcvs_mime_type:
    url = '%s/%s/%s/%s' % \
          (request.script_name, checkout_magic_path,
           os.path.dirname(request.where), url)

  url = url + '?rev=' + revision + request.amp_query
  if mime_type:
    return url + '&content-type=' + mime_type
  return url

_time_desc = {
         1 : 'second',
        60 : 'minute',
      3600 : 'hour',
     86400 : 'day',
    604800 : 'week',
   2628000 : 'month',
  31536000 : 'year',
  }

def get_time_text(request, interval, num):
  "Get some time text, possibly internationalized."
  ### some languages have even harder pluralization rules. we'll have to
  ### deal with those on demand
  if num == 0:
    return ''
  text = _time_desc[interval]
  if num == 1:
    attr = text + '_singular'
    fmt = '%d ' + text
  else:
    attr = text + '_plural'
    fmt = '%d ' + text + 's'
  try:
    fmt = getattr(request.kv.i18n.time, attr)
  except AttributeError:
    pass
  return fmt % num

def little_time(request):
  try:
    return request.kv.i18n.time.little_time
  except AttributeError:
    return 'very little time'

def html_time(request, secs, extended=0):
  secs = long(time.time()) - secs
  if secs < 2:
    return little_time(request)
  breaks = _time_desc.keys()
  breaks.sort()
  i = 0
  while i < len(breaks):
    if secs < 2 * breaks[i]:
      break
    i = i + 1
  value = breaks[i - 1]
  s = get_time_text(request, value, secs / value)

  if extended and i > 1:
    secs = secs % value
    value = breaks[i - 2]
    ext = get_time_text(request, value, secs / value)
    if ext:
      ### this is not i18n compatible. pass on it for now
      s = s + ', ' + ext
  return s

def nav_header_data(request, path, filename, rev):
  return {
    'nav_path' : clickable_path(request, path, 1, 0, 0),
    'path' : path,
    'filename' : filename,