Open Source Projects Europe

import os

import os

import shlex

import shlex

import urllib

import urllib

import sys

import sys

from uprclutils import uplog, docarturi, audiomtypes, rcldirentry, \

from uprclutils import uplog, docarturi, audiomtypes, rcldirentry, \

     rcldoctoentry, cmpentries

     rcldoctoentry, cmpentries

from recoll import recoll

from recoll import recoll

from recoll import rclconfig

from recoll import rclconfig

_foldersIdPfx = '0$uprcl$folders'

# Debug : limit processed recoll entries for speed

# Debug : limit processed recoll entries for speed

_maxrclcnt = 0

# Internal init: create the directory tree (folders view) from the doc

# Internal init: create the directory tree (folders view) from the doc

# array by splitting the url in each doc.

# array by splitting the url in each doc.

#

#

# The dirvec vector has one entry for each directory. Each entry is a

# The dirvec vector has one entry for each directory. Each entry is a

    return len(dirvec) - 1

    return len(dirvec) - 1

def _rcl2folders(docs, confdir, httphp, pathprefix):

def _rcl2folders(docs, confdir, httphp, pathprefix):

    global dirvec

    global dirvec

    dirvec = []

    dirvec = []

    rclconf = rclconfig.RclConfig(confdir)

    rclconf = rclconfig.RclConfig(confdir)

    topdirs = [os.path.expanduser(d) for d in

    topdirs = [os.path.expanduser(d) for d in

               shlex.split(rclconf.getConfParam('topdirs'))]

               shlex.split(rclconf.getConfParam('topdirs'))]

    topdirs = [d.rstrip('/') for d in topdirs]

    topdirs = [d.rstrip('/') for d in topdirs]

    if False:

    if False:

        for ent in dirvec:

        for ent in dirvec:

            uplog("%s" % ent)

            uplog("%s" % ent)

    return dirvec

    return dirvec

# Internal init: fetch all the docs by querying Recoll with [mime:*],

# Internal init: fetch all the docs by querying Recoll with [mime:*],

# which is guaranteed to match every doc without overflowing the query

# which is guaranteed to match every doc without overflowing the query

# size (because the number of mime types is limited). Something like

# size (because the number of mime types is limited). Something like

# title:* would overflow.

# title:* would overflow.

def _fetchalldocs(confdir):

def _fetchalldocs(confdir):

    allthedocs = []

    allthedocs = []

    rcldb = recoll.connect(confdir=confdir)

    rcldb = recoll.connect(confdir=confdir)

    rclq = rcldb.query()

    rclq = rcldb.query()

    rclq.execute("mime:*", stemming=0)

    rclq.execute("mime:*", stemming=0)

    while True:

    while True:

        docs = rclq.fetchmany()

        docs = rclq.fetchmany()

        for doc in docs:

        for doc in docs:

            allthedocs.append(doc)

            allthedocs.append(doc)

            totcnt += 1

            totcnt += 1

        if (_maxrclcnt > 0 and totcnt >= _maxrclcnt) or \

               len(docs) != rclq.arraysize:

               len(docs) != rclq.arraysize:

            break

            break

    uplog("Retrieved %d docs" % (totcnt,))

    return allthedocs

    return allthedocs

# Initialize (read recoll data and build tree)

def inittree(confdir, httphp, pathprefix):

def inittree(confdir, httphp, pathprefix):

    global g_alldocs, _dirvec

    g_alldocs = _fetchalldocs(confdir)

    g_alldocs = _fetchalldocs(confdir)

    _dirvec = _rcl2folders(g_alldocs, confdir, httphp, pathprefix)

    return g_alldocs

    return g_alldocs

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

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

# Browsing the initialized [folders] hierarchy

# Browsing the initialized [folders] hierarchy

# Extract dirvec index from objid, according to the way we generate them.

# Extract dirvec index from objid, according to the way we generate them.

def _objidtodiridx(pid):

def _objidtodiridx(pid):

    if not pid.startswith(_foldersIdPfx):

        raise Exception("folders.browse: bad pid %s" % pid)

        raise Exception("folders.browse: bad pid %s" % pid)

    if len(g_alldocs) == 0:

    if len(g_alldocs) == 0:

        raise Exception("folders:browse: no docs")

        raise Exception("folders:browse: no docs")

    diridx = pid[len(_foldersIdPfx):]

    if not diridx:

    if not diridx:

        diridx = 0

        diridx = 0

    else:

    else:

        if diridx[1] != 'd':

        if diridx[1] != 'd':

            raise Exception("folders:browse: called on non dir objid %s" % pid)

            raise Exception("folders:browse: called on non dir objid %s" % pid)

        diridx = int(diridx[2:])

        diridx = int(diridx[2:])

    if diridx >= len(_dirvec):

        raise Exception("folders:browse: bad pid %s" % pid)

        raise Exception("folders:browse: bad pid %s" % pid)

    return diridx

    return diridx

# Look all non-directory docs inside directory, and return the cover

# Look all non-directory docs inside directory, and return the cover

# art we find.

# art we find.

def _arturifordir(diridx):

def _arturifordir(diridx):

    for nm,ids in _dirvec[diridx].iteritems():

        if ids[1] >= 0:

        if ids[1] >= 0:

            doc = g_alldocs[ids[1]]

            doc = g_alldocs[ids[1]]

            if doc.mtype != 'inode/directory' and doc.albumarturi:

            if doc.mtype != 'inode/directory' and doc.albumarturi:

                return doc.albumarturi

                return doc.albumarturi

    entries = []

    entries = []

    # The basename call is just for diridx==0 (topdirs). Remove it if

    # The basename call is just for diridx==0 (topdirs). Remove it if

    # this proves a performance issue

    # this proves a performance issue

    for nm,ids in _dirvec[diridx].iteritems():

        if nm == "..":

        if nm == "..":

            continue

            continue

        thisdiridx = ids[0]

        thisdiridx = ids[0]

        thisdocidx = ids[1]

        thisdocidx = ids[1]

        if thisdocidx >= 0:

        if thisdocidx >= 0:

        if thisdiridx >= 0:

        if thisdiridx >= 0:

            # Skip empty directories

            # Skip empty directories

            if len(dirvec[thisdiridx]) == 1:

            if len(dirvec[thisdiridx]) == 1:

                continue

                continue

            id = _foldersIdPfx + '$' + 'd' + str(thisdiridx)

            if doc and doc.albumarturi:

            if doc and doc.albumarturi:

                arturi = doc.albumarturi

                arturi = doc.albumarturi

            else:

            else:

                arturi = _arturifordir(thisdiridx)

                arturi = _arturifordir(thisdiridx)

            entries.append(rcldirentry(id, pid, os.path.basename(nm),

            entries.append(rcldirentry(id, pid, os.path.basename(nm),

            # Not a directory. docidx had better been set

            # Not a directory. docidx had better been set

            if thisdocidx == -1:

            if thisdocidx == -1:

                uplog("folders:docidx -1 for non-dir entry %s"%nm)

                uplog("folders:docidx -1 for non-dir entry %s"%nm)

                continue

                continue

            doc = g_alldocs[thisdocidx]

            doc = g_alldocs[thisdocidx]

            id = _foldersIdPfx + '$i' + str(thisdocidx)

            e = rcldoctoentry(id, pid, httphp, pathprefix, doc)

            e = rcldoctoentry(id, pid, httphp, pathprefix, doc)

            if e:

            if e:

                entries.append(e)

                entries.append(e)

    return sorted(entries, cmp=cmpentries)

    return sorted(entries, cmp=cmpentries)

    if diridx == 0:

    if diridx == 0:

        return "/"

        return "/"

    lpath = []

    lpath = []

    while True:

    while True:

        fathidx = _dirvec[diridx][".."][0]

        for nm, ids in _dirvec[fathidx].iteritems():

            if ids[0] == diridx:

            if ids[0] == diridx:

                lpath.append(nm)

                lpath.append(nm)

                break

                break

        diridx = fathidx

        diridx = fathidx

        if diridx == 0:

        if diridx == 0: