Stop showing symlinks in search results, closes #458

dxr/app.py

@@ -209,15 +209,18 @@ def browse(tree, path=''):
         return _browse_folder(tree, path.rstrip('/'), config)
     except NotFound:
         frozen = frozen_config(tree)
-        # Grab the FILE doc, just for the sidebar nav links:
+        # Grab the FILE doc, just for the sidebar nav links and the symlink target:
         files = filtered_query(
             frozen['es_alias'],
             FILE,
             filter={'path': path},
             size=1,
-            include=['links'])
+            include=['link', 'links'])
         if not files:
             raise NotFound
+        if 'link' in files[0]:
+            # Then this path is a symlink, so redirect to the real thing.
+            return redirect(url_for('.browse', tree=tree, path=files[0]['link'][0]))
 
         lines = filtered_query(
             frozen['es_alias'],
@@ -279,7 +282,7 @@ def _browse_folder(tree, path, config):
              f['name'],
              decode_es_datetime(f['modified']) if 'modified' in f else None,
              f.get('size'),
-             url_for('.browse', tree=tree, path=f['path'][0]),
+             url_for('.browse', tree=tree, path=f.get('link', f['path'])[0]),
              f.get('is_binary', [False])[0])
             for f in files_and_folders])
 

dxr/build.py

@@ -457,7 +457,10 @@ def index_file(tree, tree_indexers, path, es, index):
 
     rel_path = relpath(path, tree.source_folder)
     is_text = isinstance(contents, unicode)
-    if is_text:
+    is_link = islink(path)
+    # Index by line if the contents are text and the path is not a symlink.
+    index_by_line = is_text and not is_link
+    if index_by_line:
         lines = contents.splitlines(True)
         num_lines = len(lines)
         needles_by_line = [{} for _ in xrange(num_lines)]
@@ -471,10 +474,11 @@ def index_file(tree, tree_indexers, path, es, index):
         if file_to_index.is_interesting():
             # Per-file stuff:
             append_update(needles, file_to_index.needles())
-            linkses.append(file_to_index.links())
+            if not is_link:
+                linkses.append(file_to_index.links())
 
             # Per-line stuff:
-            if is_text:
+            if index_by_line:
                 refses.append(file_to_index.refs())
                 regionses.append(file_to_index.regions())
                 append_update_by_line(needles_by_line,
@@ -514,9 +518,8 @@ def docs():
             doc['links'] = links
         yield es.index_op(doc, doc_type=FILE)
 
-        # Index all the lines. If it's an empty file (no lines), don't bother
-        # ES. It hates empty dicts.
-        if is_text and needles_by_line:
+        # Index all the lines.
+        if index_by_line:
             for total, annotations_for_this_line, tags in izip(
                     needles_by_line,
                     annotations_by_line,

dxr/indexers.py

@@ -4,7 +4,7 @@
 from collections import namedtuple
 import json
 from operator import itemgetter
-from os.path import join
+from os.path import join, islink
 from warnings import warn
 
 from funcy import group_by, decorator, imapcat
@@ -209,10 +209,13 @@ def is_interesting(self):
         :meth:`~dxr.indexers.FileToSkim.links()`,
         :meth:`~dxr.indexers.FileToSkim.refs()`, etc.
 
-        The default implementation selects only text files.
+        The default implementation selects only text files that are not symlinks.
+        Note: even if a plugin decides that symlinks are interesting, it should
+        remember that links, refs, regions and by-line annotations will not be
+        called because views of symlinks redirect to the original file.
 
         """
-        return self.contains_text()
+        return self.contains_text() and not self.is_link()
 
     def links(self):
         """Return an iterable of links for the navigation pane::
@@ -306,6 +309,15 @@ def absolute_path(self):
         """
         return join(self.tree.source_folder, self.path)
 
+    def is_link(self):
+        """Return whether the file is a symlink.
+
+        Note: symlinks are never displayed in file browsing; a request for a symlink redirects
+        to its target.
+
+        """
+        return islink(self.absolute_path())
+
     # Private methods:
 
     def _line_offsets(self):
@@ -393,6 +405,10 @@ def needles_by_line(self):
         may be dicts, in which case common keys get merged by
         :func:`~dxr.utils.append_update()`.
 
+        This method is not called on symlink files, to maintain the illusion
+        that they do not have contents, seeing as they cannot be viewed in
+        file browsing.
+
         """
         return []
 

dxr/plugins/core.py

@@ -2,7 +2,7 @@
 
 from base64 import b64encode
 from itertools import chain
-from os.path import relpath, splitext
+from os.path import relpath, splitext, islink, realpath
 import re
 
 from flask import url_for
@@ -60,6 +60,11 @@
 
             'ext': EXT_MAPPING,
 
+            'link': {  # the target path if this FILE is a symlink
+                'type': 'string',
+                'index': 'not_analyzed'
+            },
+
             # Folder listings query by folder and then display filename, size,
             # and mod date.
             'folder': {  # path/to/a/folder
@@ -433,6 +438,9 @@ def __init__(self, path, contents, plugin_name, tree, vcs):
 
     def needles(self):
         """Fill out path (and path.trigrams)."""
+        if self.is_link():
+            # realpath will keep following symlinks until it gets to the 'real' thing.
+            yield 'link', relpath(realpath(self.absolute_path()), self.tree.source_folder)
         yield 'path', self.path
         extension = splitext(self.path)[1]
         if extension:

dxr/plugins/python/indexers.py

@@ -1,7 +1,8 @@
 import ast
 import token
 import tokenize
-from StringIO import StringIO
+from os.path import islink
+from cStringIO import StringIO
 
 from dxr.build import unignored
 from dxr.filters import FILE, LINE
@@ -301,4 +302,4 @@ def is_interesting(path):
     analyze.
 
     """
-    return path.endswith('.py')
+    return path.endswith('.py') and not islink(path)

dxr/query.py

@@ -140,6 +140,8 @@ def group_filters_by_name(predicate):
             # Don't show folders yet in search results. I don't think the JS
             # is able to handle them.
             ors.append({'term': {'is_folder': False}})
+            # Filter out all FILE docs who are links.
+            ors.append({'not': {'exists': {'field': 'link'}}})
 
         if ors:
             query = {

tests/test_symlink/code/README.mkd

@@ -0,0 +1 @@
+This file should be indexed happily.

tests/test_symlink/code/link.mkd

@@ -0,0 +1 @@
+README.mkd

tests/test_symlink/dxr.config

@@ -0,0 +1,13 @@
+[DXR]
+enabled_plugins     = pygmentize clang python
+es_index            = dxr_test_{format}_{tree}_{unique}
+es_alias            = dxr_test_{format}_{tree}
+es_catalog_index    = dxr_test_catalog
+
+[code]
+source_folder       = code
+build_command       =
+clean_command       =
+
+    [[python]]
+    python_path     = ./

tests/test_symlink/test_symlink.py

@@ -0,0 +1,36 @@
+from nose.tools import eq_, ok_
+
+from dxr.testing import DxrInstanceTestCase
+
+
+class SymlinkTests(DxrInstanceTestCase):
+    def test_follows_link(self):
+        """Test that a symlink listed in a browsing view will actually point to the real target.
+
+        """
+        response = self.client().get('/code/source/')
+        # Make sure the browse view actually shows the symlink...
+        ok_('link.mkd' in response.data)
+        # ...but links to the real file instead.
+        ok_('<a href="/code/source/link.mkd"' not in response.data)
+
+    def test_file_search(self):
+        """Make sure that searching for path:<symlink name> does not return the symlink.
+
+        """
+        self.found_files_eq('path:mkd', ['README.mkd'])
+
+    def test_line_search(self):
+        """Make sure that searching for contents within the real file does not return duplicates
+        in the symlink.
+
+        """
+        self.found_files_eq('happily', ['README.mkd'])
+
+    def test_redirect(self):
+        """Make sure that a direct link to a symlink redirects to the real file.
+
+        """
+        response = self.client().get('/code/source/link.mkd')
+        eq_(response.status_code, 302)
+        ok_(response.headers['Location'].endswith('/code/source/README.mkd'))