apache.py

Mod_python is an Apache module that embeds the Python interpreter within the server. With mod_python

                        # result.

                        if len(value.args) == 2:
                            (result, status) = value.args
                            if status:
                                req.status = status
                        else:
                            result = value.args[0]

                    assert (type(result) == types.IntType), \
                            _result_warning % type(result)

                    # stop cycling through handlers
                    if result not in expected:
                        break

                elif hlist.silent:
                    # A missing handler when in silent mode will
                    # only propagate DECLINED if it is the first
                    # and only handler.

                    if result == HTTP_INTERNAL_SERVER_ERROR:
                        result = DECLINED

                hlist.next()

        except PROG_TRACEBACK, traceblock:
            # Program run-time error
            try:
                (etype, value, traceback) = traceblock
                result = self.ReportError(etype, value, traceback, req=req,
                                          phase=req.phase, hname=hlist.handler,
                                          debug=debug)
            finally:
                traceback = None

        except:
            # Any other rerror (usually parsing)
            try:
                exc_type, exc_value, exc_traceback = sys.exc_info()
                result = self.ReportError(exc_type, exc_value, exc_traceback, req=req,
                                          phase=req.phase, hname=hlist.handler, debug=debug)
            finally:
                exc_traceback = None

        return result

    def IncludeDispatch(self, filter, tag, code):

        try:
            config = filter.req.get_config()
            debug = int(config.get("PythonDebug", 0))

            if not hasattr(filter.req,"ssi_globals"):
                filter.req.ssi_globals = {}

            filter.req.ssi_globals["filter"] = filter

            filter.req.ssi_globals["__file__"] = filter.req.filename

            code = code.replace('\r\n', '\n').rstrip()

            if tag == 'eval':
                result = eval(code, filter.req.ssi_globals)
                if result is not None:
                    filter.write(str(result))
            elif tag == 'exec':
                exec(code, filter.req.ssi_globals)

            filter.flush()

        except:
            try:
                exc_type, exc_value, exc_traceback = sys.exc_info()
                result = self.ReportError(exc_type, exc_value, exc_traceback,
                                          filter=filter, phase=filter.name,
                                          hname=filter.req.filename,
                                          debug=debug)
            finally:
                exc_traceback = None

            raise

        filter.req.ssi_globals["filter"] = None

        return OK

    def ImportDispatch(self, name):

        config = main_server.get_config()

        debug = int(config.get("PythonDebug", "0"))

        # evaluate pythonpath and set sys.path to
        # resulting value if not already done

        if config.has_key("PythonPath"): 
            _path_cache_lock.acquire() 
            try: 
                pathstring = config["PythonPath"] 
                if not _path_cache.has_key(pathstring): 
                    newpath = eval(pathstring) 
                    _path_cache[pathstring] = None 
                    sys.path[:] = newpath 
            finally: 
                _path_cache_lock.release()

        # split module::function
        l = name.split('::', 1)
        module_name = l[0]
        func_name = None
        if len(l) != 1:
            func_name = l[1]

        module = import_module(module_name, log=debug)

        if func_name:
            getattr(module, func_name)()

    def ReportError(self, etype, evalue, etb, req=None, filter=None, srv=None,
                    phase="N/A", hname="N/A", debug=0):
        """
        This function is only used when debugging is on.
        It sends the output similar to what you'd see
        when using Python interactively to the browser
        """
        try:       # try/finally
            try:        # try/except

                if str(etype) == "exceptions.IOError" \
                   and str(evalue)[:5] == "Write":
                    # if this is an IOError while writing to client,
                    # it is probably better not to try to write to the cleint
                    # even if debug is on.
                    debug = 0

                # write to log
                for e in traceback.format_exception(etype, evalue, etb):
                    s = "%s %s: %s" % (phase, hname, e[:-1])
                    if req:
                        req.log_error(s, APLOG_NOERRNO|APLOG_ERR)
                    else:
                        _apache.log_error(s, APLOG_NOERRNO|APLOG_ERR, srv)

                if not debug or not req:
                    return HTTP_INTERNAL_SERVER_ERROR
                else:
                    # write to client
                    req.status = HTTP_INTERNAL_SERVER_ERROR
                    req.content_type = 'text/html'

                    s = '\n<pre>\nMod_python error: "%s %s"\n\n' % (phase, hname)
                    for e in traceback.format_exception(etype, evalue, etb):
                        s = s + cgi.escape(e) + '\n'
                    s = s + "</pre>\n"

                    if filter:
                        filter.write(s)
                        filter.flush()
                    else:
                        req.write(s)

                    return DONE
            except:
                # last try
                traceback.print_exc()
                sys.stderr.flush()

        finally:
            # erase the traceback
            etb = None
            # we do not return anything

def import_module(module_name, autoreload=1, log=0, path=None):
    """
    Get the module to handle the request. If
    autoreload is on, then the module will be reloaded
    if it has changed since the last import.
    """

    # nlehuen: this is a big lock, we'll have to refine it later to get better performance.
    # For now, we'll concentrate on thread-safety.
    imp.acquire_lock()
    try:
        # (Re)import
        if sys.modules.has_key(module_name):

            # The module has been imported already
            module = sys.modules[module_name]

            # but is it in the path?
            file = module.__dict__.get("__file__")

            # the "and not" part of this condition is to prevent execution
            # of arbitrary already imported modules, such as os. The
            # reason we use startswith as opposed to exact match is that
            # modules inside packages are actually in subdirectories.

            if not file or (path and not filter(file.startswith, path)):
                # there is a script by this name already imported, but it's in
                # a different directory, therefore it's a different script
                mtime, oldmtime = 0, -1
            elif autoreload:
                oldmtime = module.__dict__.get("__mtime__", 0)
                mtime = module_mtime(module)
            else:
                mtime, oldmtime = 0, 0

        else:
            mtime, oldmtime = 0, -1

        if mtime != oldmtime:

            # Import the module
            if log:
                if path:
                    s = "mod_python: (Re)importing module '%s' with path set to '%s'" % (module_name, path)
                else:
                    s = "mod_python: (Re)importing module '%s'" % module_name
                _apache.log_error(s, APLOG_NOERRNO|APLOG_NOTICE)

            parent = None
            parts = module_name.split('.')
            for i in range(len(parts)):
                f, p, d = imp.find_module(parts[i], path)
                try:
                    mname = ".".join(parts[:i+1])
                    module = imp.load_module(mname, f, p, d)
                    if parent:
                        setattr(parent,parts[i],module)
                    parent = module
                finally:
                    if f: f.close()
                if hasattr(module, "__path__"):
                    path = module.__path__

            if mtime == 0:
                mtime = module_mtime(module)

            module.__mtime__ = mtime

        return module
    finally:
        imp.release_lock()

def module_mtime(module):
    """Get modification time of module"""
    mtime = 0
    if module.__dict__.has_key("__file__"):

        filepath = module.__file__

        try:
            # this try/except block is a workaround for a Python bug in
            # 2.0, 2.1 and 2.1.1. See
            # http://sourceforge.net/tracker/?group_id=5470&atid=105470&func=detail&aid=422004

            if os.path.exists(filepath):
                mtime = os.path.getmtime(filepath)

            if os.path.exists(filepath[:-1]) :
                mtime = max(mtime, os.path.getmtime(filepath[:-1]))

        except OSError: pass

    return mtime

def resolve_object(module, object_str, arg=None, silent=0):
    """
    This function traverses the objects separated by .
    (period) to find the last one we're looking for:

       From left to right, find objects, if it is
       an unbound method of a class, instantiate the
       class passing the request as single argument

    'arg' is sometimes req, sometimes filter,
    sometimes connection
    """

    obj = module

    for obj_str in  object_str.split('.'):

        parent = obj

        # don't throw attribute errors when silent
        if silent and not hasattr(obj, obj_str):
            return None

        # this adds a little clarity if we have an attriute error
        if obj == module and not hasattr(module, obj_str):
            if hasattr(module, "__file__"):
                s = "module '%s' contains no '%s'" % (module.__file__, obj_str)
                raise AttributeError, s

        obj = getattr(obj, obj_str)

        if hasattr(obj, "im_self") and not obj.im_self:
            # this is an unbound method, its class
            # needs to be instantiated
            instance = parent(arg)
            obj = getattr(instance, obj_str)

    return obj

def build_cgi_env(req):
    """
    Utility function that returns a dictionary of
    CGI environment variables as described in
    http://hoohoo.ncsa.uiuc.edu/cgi/env.html
    """

    req.add_common_vars()
    env = req.subprocess_env.copy()

    if req.path_info and len(req.path_info) > 0:
        env["SCRIPT_NAME"] = req.uri[:-len(req.path_info)]
    else:
        env["SCRIPT_NAME"] = req.uri

    env["GATEWAY_INTERFACE"] = "Python-CGI/1.1"

    # you may want to comment this out for better security
    if req.headers_in.has_key("authorization"):
        env["HTTP_AUTHORIZATION"] = req.headers_in["authorization"]

    return env

class NullIO:
    """ Abstract IO
    """
    def tell(self): return 0
    def read(self, n = -1): return ""
    def readline(self, length = None): return ""
    def readlines(self): return []
    def write(self, s): pass
    def writelines(self, list):
        self.write("".join(list))
    def isatty(self): return 0
    def flush(self): pass
    def close(self): pass
    def seek(self, pos, mode = 0): pass

class CGIStdin(NullIO):

    def __init__(self, req):
        self.pos = 0
        self.req = req
        self.BLOCK = 65536 # 64K
        # note that self.buf sometimes contains leftovers
        # that were read, but not used when readline was used
        self.buf = ""

    def read(self, n = -1):
        if n == 0:
            return ""
        if n == -1:
            s = self.req.read(self.BLOCK)
            while s:
                self.buf = self.buf + s
                self.pos = self.pos + len(s)
                s = self.req.read(self.BLOCK)
            result = self.buf
            self.buf = ""
            return result
        else:
            if self.buf:
                s = self.buf[:n]
                n = n - len(s)
            else:
                s = ""
            s = s + self.req.read(n)
            self.pos = self.pos + len(s)
            return s

    def readlines(self):
        s = (self.buf + self.read()).split('\n')
        return map(lambda s: s + '\n', s)

    def readline(self, n = -1):

        if n == 0:
            return ""

        # fill up the buffer
        self.buf = self.buf + self.req.read(self.BLOCK)

        # look for \n in the buffer
        i = self.buf.find('\n')
        while i == -1: # if \n not found - read more
            if (n != -1) and (len(self.buf) >= n): # we're past n
                i = n - 1
                break
            x = len(self.buf)
            self.buf = self.buf + self.req.read(self.BLOCK)
            if len(self.buf) == x: # nothing read, eof
                i = x - 1
                break
            i = self.buf.find('\n', x)