Add APIs for git attributes

This adds APIs for querying git attributes.  In addition to
the new API in include/git2/attr.h, most of the action is in
src/attr_file.[hc] which contains utilities for dealing with
a single attributes file, and src/attr.[hc] which contains
the implementation of the APIs that merge all applicable
attributes files.

12 files changed, 911 insertions, 14 deletions

diff --git a/src/attr.c b/src/attr.c
new file mode 100644
index 000000000..d8e7095b1
--- /dev/null
+++ b/src/attr.c
@@ -0,0 +1,311 @@
+#include "attr.h"
+#include "buffer.h"
+#include "fileops.h"
+#include "config.h"
+#include <ctype.h>
+
+#define GIT_ATTR_FILE_INREPO	"info/attributes"
+#define GIT_ATTR_FILE			".gitattributes"
+#define GIT_ATTR_FILE_SYSTEM	"/etc/gitattributes"
+#if GIT_WIN32
+#define GIT_ATTR_FILE_WIN32	L"%PROGRAMFILES%\\Git\\etc\\gitattributes"
+#endif
+
+static int collect_attr_files(
+	git_repository *repo, const char *path, git_vector *files);
+
+
+int git_attr_get(
+    git_repository *repo, const char *pathname,
+	const char *name, const char **value)
+{
+	int error;
+	git_attr_path path;
+	git_vector files = GIT_VECTOR_INIT;
+	unsigned int i, j;
+	git_attr_file *file;
+	git_attr_name attr;
+	git_attr_rule *rule;
+
+	*value = NULL;
+
+	if ((error = git_attr_path__init(&path, pathname)) < GIT_SUCCESS ||
+		(error = collect_attr_files(repo, pathname, &files)) < GIT_SUCCESS)
+		return git__rethrow(error, "Could not get attribute for %s", pathname);
+
+	attr.name = name;
+	attr.name_hash = git_attr_file__name_hash(name);
+
+	git_vector_foreach(&files, i, file) {
+
+		git_attr_file__foreach_matching_rule(file, &path, j, rule) {
+			int pos = git_vector_bsearch(&rule->assigns, &attr);
+			git_clearerror(); /* okay if search failed */
+
+			if (pos >= 0) {
+				*value = ((git_attr_assignment *)git_vector_get(
+							  &rule->assigns, pos))->value;
+				goto found;
+			}
+		}
+	}
+
+found:
+	git_vector_free(&files);
+
+	return error;
+}
+
+
+typedef struct {
+	git_attr_name name;
+	git_attr_assignment *found;
+} attr_get_many_info;
+
+int git_attr_get_many(
+    git_repository *repo, const char *pathname,
+    size_t num_attr, const char **names, const char **values)
+{
+	int error;
+	git_attr_path path;
+	git_vector files = GIT_VECTOR_INIT;
+	unsigned int i, j, k;
+	git_attr_file *file;
+	git_attr_rule *rule;
+	attr_get_many_info *info = NULL;
+	size_t num_found = 0;
+
+	memset(values, 0, sizeof(const char *) * num_attr);
+
+	if ((error = git_attr_path__init(&path, pathname)) < GIT_SUCCESS ||
+		(error = collect_attr_files(repo, pathname, &files)) < GIT_SUCCESS)
+		return git__rethrow(error, "Could not get attributes for %s", pathname);
+
+	if ((info = git__calloc(num_attr, sizeof(attr_get_many_info))) == NULL) {
+		git__rethrow(GIT_ENOMEM, "Could not get attributes for %s", pathname);
+		goto cleanup;
+	}
+
+	git_vector_foreach(&files, i, file) {
+
+		git_attr_file__foreach_matching_rule(file, &path, j, rule) {
+
+			for (k = 0; k < num_attr; k++) {
+				int pos;
+
+				if (info[k].found != NULL) /* already found assignment */
+					continue;
+
+				if (!info[k].name.name) {
+					info[k].name.name = names[k];
+					info[k].name.name_hash = git_attr_file__name_hash(names[k]);
+				}
+
+				pos = git_vector_bsearch(&rule->assigns, &info[k].name);
+				git_clearerror(); /* okay if search failed */
+
+				if (pos >= 0) {
+					info[k].found = (git_attr_assignment *)
+						git_vector_get(&rule->assigns, pos);
+					values[k] = info[k].found->value;
+
+					if (++num_found == num_attr)
+						goto cleanup;
+				}
+			}
+		}
+	}
+
+cleanup:
+	git_vector_free(&files);
+	git__free(info);
+
+	return error;
+}
+
+
+int git_attr_foreach(
+    git_repository *repo, const char *pathname,
+	int (*callback)(const char *name, const char *value, void *payload),
+	void *payload)
+{
+	int error;
+	git_attr_path path;
+	git_vector files = GIT_VECTOR_INIT;
+	unsigned int i, j, k;
+	git_attr_file *file;
+	git_attr_rule *rule;
+	git_attr_assignment *assign;
+	git_hashtable *seen = NULL;
+
+	if ((error = git_attr_path__init(&path, pathname)) < GIT_SUCCESS ||
+		(error = collect_attr_files(repo, pathname, &files)) < GIT_SUCCESS)
+		return git__rethrow(error, "Could not get attributes for %s", pathname);
+
+	seen = git_hashtable_alloc(8, git_hash__strhash_cb, git_hash__strcmp_cb);
+	if (!seen) {
+		error = GIT_ENOMEM;
+		goto cleanup;
+	}
+
+	git_vector_foreach(&files, i, file) {
+
+		git_attr_file__foreach_matching_rule(file, &path, j, rule) {
+
+			git_vector_foreach(&rule->assigns, k, assign) {
+				/* skip if higher priority assignment was already seen */
+				if (git_hashtable_lookup(seen, assign->name))
+					continue;
+
+				error = git_hashtable_insert(seen, assign->name, assign);
+				if (error != GIT_SUCCESS)
+					goto cleanup;
+
+				error = callback(assign->name, assign->value, payload);
+				if (error != GIT_SUCCESS)
+					goto cleanup;
+			}
+		}
+	}
+
+cleanup:
+	if (seen)
+		git_hashtable_free(seen);
+	git_vector_free(&files);
+
+	if (error != GIT_SUCCESS)
+		(void)git__rethrow(error, "Could not get attributes for %s", pathname);
+
+	return error;
+}
+
+
+/* add git_attr_file to vector of files, loading if needed */
+static int push_attrs(
+	git_repository *repo,
+	git_vector     *files,
+	const char     *base,
+	const char     *filename)
+{
+	int error = GIT_SUCCESS;
+	git_attr_cache *cache = &repo->attrcache;
+	git_buf path = GIT_BUF_INIT;
+	git_attr_file *file;
+	int add_to_cache = 0;
+
+	if (cache->files == NULL) {
+		cache->files = git_hashtable_alloc(
+			8, git_hash__strhash_cb, git_hash__strcmp_cb);
+		if (!cache->files)
+			return git__throw(GIT_ENOMEM, "Could not create attribute cache");
+	}
+
+	if ((error = git_path_prettify(&path, filename, base)) < GIT_SUCCESS) {
+		if (error == GIT_EOSERR)
+			/* file was not found -- ignore error */
+			error = GIT_SUCCESS;
+		goto cleanup;
+	}
+
+	/* either get attr_file from cache or read from disk */
+	file = git_hashtable_lookup(cache->files, path.ptr);
+	if (file == NULL) {
+		error = git_attr_file__from_file(&file, path.ptr);
+		add_to_cache = (error == GIT_SUCCESS);
+	}
+
+	if (file != NULL) {
+		/* add file to vector, if we found it */
+		error = git_vector_insert(files, file);
+
+		/* add file to cache, if it is new */
+		/* do this after above step b/c it is not critical */
+		if (error == GIT_SUCCESS && add_to_cache && file->path != NULL)
+			error = git_hashtable_insert(cache->files, file->path, file);
+	}
+
+cleanup:
+	git_buf_free(&path);
+	return error;
+}
+
+
+static int collect_attr_files(
+	git_repository *repo, const char *path, git_vector *files)
+{
+	int error = GIT_SUCCESS;
+	git_buf dir = GIT_BUF_INIT;
+	git_config *cfg;
+	const char *workdir = git_repository_workdir(repo);
+
+	if ((error = git_vector_init(files, 4, NULL)) < GIT_SUCCESS)
+		goto cleanup;
+
+	if ((error = git_path_prettify(&dir, path, workdir)) < GIT_SUCCESS)
+		goto cleanup;
+
+	if (git_futils_isdir(dir.ptr) != GIT_SUCCESS) {
+		git_path_dirname_r(&dir, dir.ptr);
+		git_path_to_dir(&dir);
+		if ((error = git_buf_lasterror(&dir)) < GIT_SUCCESS)
+			goto cleanup;
+	}
+
+	/* in precendence order highest to lowest:
+	 * - $GIT_DIR/info/attributes
+	 * - path components with .gitattributes
+	 * - config core.attributesfile
+	 * - $GIT_PREFIX/etc/gitattributes
+	 */
+
+	error = push_attrs(repo, files, repo->path_repository, GIT_ATTR_FILE_INREPO);
+	if (error < GIT_SUCCESS)
+		goto cleanup;
+
+	if (workdir && git__prefixcmp(dir.ptr, workdir) == 0) {
+		ssize_t rootlen = (ssize_t)strlen(workdir);
+
+		do {
+			error = push_attrs(repo, files, dir.ptr, GIT_ATTR_FILE);
+			if (error == GIT_SUCCESS) {
+				git_path_dirname_r(&dir, dir.ptr);
+				git_path_to_dir(&dir);
+				error = git_buf_lasterror(&dir);
+			}
+		} while (!error && dir.size >= rootlen);
+	} else {
+		error = push_attrs(repo, files, dir.ptr, GIT_ATTR_FILE);
+	}
+	if (error < GIT_SUCCESS)
+		goto cleanup;
+
+	if (git_repository_config(&cfg, repo) == GIT_SUCCESS) {
+		const char *core_attribs = NULL;
+		git_config_get_string(cfg, "core.attributesfile", &core_attribs);
+		git_clearerror(); /* don't care if attributesfile is not set */
+		if (core_attribs)
+			error = push_attrs(repo, files, NULL, core_attribs);
+		git_config_free(cfg);
+	}
+
+	if (error == GIT_SUCCESS)
+		error = push_attrs(repo, files, NULL, GIT_ATTR_FILE_SYSTEM);
+
+ cleanup:
+	if (error < GIT_SUCCESS) {
+		git__rethrow(error, "Could not get attributes for '%s'", path);
+		git_vector_free(files);
+	}
+	git_buf_free(&dir);
+
+	return error;
+}
+
+
+void git_repository__attr_cache_free(git_attr_cache *attrs)
+{
+	if (attrs && attrs->files) {
+		git_hashtable_free(attrs->files);
+		attrs->files = NULL;
+	}
+}
diff --git a/src/attr.h b/src/attr.h
new file mode 100644
index 000000000..518fb9d3b
--- /dev/null
+++ b/src/attr.h
@@ -0,0 +1,21 @@
+/*
+ * Copyright (C) 2009-2011 the libgit2 contributors
+ *
+ * This file is part of libgit2, distributed under the GNU GPL v2 with
+ * a Linking Exception. For full terms see the included COPYING file.
+ */
+#ifndef INCLUDE_attr_h__
+#define INCLUDE_attr_h__
+
+#include "hashtable.h"
+#include "attr_file.h"
+
+/* EXPORT */
+typedef struct {
+	git_hashtable *files;		/* hash path to git_attr_file */
+} git_attr_cache;
+
+extern void git_repository__attr_cache_free(git_attr_cache *attrs);
+
+#endif
+
diff --git a/src/attr_file.c b/src/attr_file.c
new file mode 100644
index 000000000..5d159db00
--- /dev/null
+++ b/src/attr_file.c
@@ -0,0 +1,456 @@
+#include "common.h"
+#include "attr_file.h"
+#include "filebuf.h"
+#include <ctype.h>
+
+const char *git_attr__true  = "[internal]__TRUE__";
+const char *git_attr__false = "[internal]__FALSE__";
+
+static int parse_fnmatch(git_attr_fnmatch *spec, const char **base);
+static int parse_assigns(git_vector *assigns, const char **base);
+static int free_rule(git_attr_rule *rule);
+static int sort_by_hash_and_name(const void *a_raw, const void *b_raw);
+
+int git_attr_file__from_buffer(git_attr_file **out, const char *buffer)
+{
+	int error = GIT_SUCCESS;
+	git_attr_file *attrs = NULL;
+	const char *scan = NULL;
+	git_attr_rule *rule = NULL;
+
+	*out = NULL;
+
+	attrs = git__calloc(1, sizeof(git_attr_file));
+	if (attrs == NULL)
+		return git__throw(GIT_ENOMEM, "Could not allocate attribute storage");
+
+	attrs->path = NULL;
+
+	error = git_vector_init(&attrs->rules, 4, NULL);
+	if (error != GIT_SUCCESS) {
+		git__rethrow(error, "Could not initialize attribute storage");
+		goto cleanup;
+	}
+
+	scan = buffer;
+
+	while (error == GIT_SUCCESS && *scan) {
+		/* allocate rule if needed */
+		if (!rule && !(rule = git__calloc(1, sizeof(git_attr_rule)))) {
+			error = GIT_ENOMEM;
+			break;
+		}
+
+		/* parse the next "pattern attr attr attr" line */
+		if (!(error = parse_fnmatch(&rule->match, &scan)) &&
+			!(error = parse_assigns(&rule->assigns, &scan)))
+			error = git_vector_insert(&attrs->rules, rule);
+
+		/* if the rule wasn't a pattern, on to the next */
+		if (error != GIT_SUCCESS) {
+			free_rule(rule); /* release anything partially allocated */
+			if (error == GIT_ENOTFOUND)
+				error = GIT_SUCCESS;
+		} else {
+			rule = NULL; /* vector now "owns" the rule */
+		}
+	}
+
+cleanup:
+	if (error != GIT_SUCCESS) {
+		git_attr_file__free(attrs);
+		git__free(attrs);
+	} else {
+		*out = attrs;
+	}
+
+	return error;
+}
+
+int git_attr_file__from_file(git_attr_file **out, const char *path)
+{
+	int error = GIT_SUCCESS;
+	git_fbuffer fbuf = GIT_FBUFFER_INIT;
+
+	*out = NULL;
+
+	if ((error = git_futils_readbuffer(&fbuf, path)) < GIT_SUCCESS ||
+		(error = git_attr_file__from_buffer(out, fbuf.data)) < GIT_SUCCESS)
+	{
+		git__rethrow(error, "Could not open attribute file '%s'", path);
+	} else {
+		/* save path (okay to fail) */
+		(*out)->path = git__strdup(path);
+	}
+
+	git_futils_freebuffer(&fbuf);
+
+	return error;
+}
+
+void git_attr_file__free(git_attr_file *file)
+{
+	unsigned int i;
+	git_attr_rule *rule;
+
+	if (!file)
+		return;
+
+	git_vector_foreach(&file->rules, i, rule) {
+		free_rule(rule);
+	}
+
+	git_vector_free(&file->rules);
+
+	git__free(file->path);
+	file->path = NULL;
+}
+
+unsigned long git_attr_file__name_hash(const char *name)
+{
+	unsigned long h = 5381;
+	int c;
+	assert(name);
+	while ((c = (int)*name++) != 0)
+		h = ((h << 5) + h) + c;
+	return h;
+}
+
+
+int git_attr_file__lookup_one(
+	git_attr_file *file,
+	const git_attr_path *path,
+	const char *attr,
+	const char **value)
+{
+	unsigned int i;
+	git_attr_name name;
+	git_attr_rule *rule;
+
+	*value = NULL;
+
+	name.name = attr;
+	name.name_hash = git_attr_file__name_hash(attr);
+
+	git_attr_file__foreach_matching_rule(file, path, i, rule) {
+		int pos = git_vector_bsearch(&rule->assigns, &name);
+		git_clearerror(); /* okay if search failed */
+
+		if (pos >= 0) {
+			*value = ((git_attr_assignment *)
+					  git_vector_get(&rule->assigns, pos))->value;
+			break;
+		}
+	}
+
+	return GIT_SUCCESS;
+}
+
+
+int git_attr_rule__match_path(
+	git_attr_rule *rule,
+	const git_attr_path *path)
+{
+	int matched = FNM_NOMATCH;
+
+	if (rule->match.directory && !path->is_dir)
+		return matched;
+
+	if (rule->match.fullpath)
+		matched = p_fnmatch(rule->match.pattern, path->path, FNM_PATHNAME);
+	else
+		matched = p_fnmatch(rule->match.pattern, path->basename, 0);
+
+	if (rule->match.negative)
+		matched = (matched == GIT_SUCCESS) ? FNM_NOMATCH : GIT_SUCCESS;
+
+	return matched;
+}
+
+git_attr_assignment *git_attr_rule__lookup_assignment(
+	git_attr_rule *rule, const char *name)
+{
+	int pos;
+	git_attr_name key;
+	key.name = name;
+	key.name_hash = git_attr_file__name_hash(name);
+
+	pos = git_vector_bsearch(&rule->assigns, &key);
+	git_clearerror(); /* okay if search failed */
+
+	return (pos >= 0) ? git_vector_get(&rule->assigns, pos) : NULL;
+}
+
+int git_attr_path__init(
+	git_attr_path *info, const char *path)
+{
+	info->path = path;
+	info->basename = strrchr(path, '/');
+	if (info->basename)
+		info->basename++;
+	if (!info->basename || !*info->basename)
+		info->basename = path;
+	info->is_dir = (git_futils_isdir(path) == GIT_SUCCESS);
+	return GIT_SUCCESS;
+}
+
+
+/*
+ * From gitattributes(5):
+ *
+ * Patterns have the following format:
+ *
+ * - A blank line matches no files, so it can serve as a separator for
+ *   readability.
+ *
+ * - A line starting with # serves as a comment.
+ *
+ * - An optional prefix ! which negates the pattern; any matching file
+ *   excluded by a previous pattern will become included again. If a negated
+ *   pattern matches, this will override lower precedence patterns sources.
+ *
+ * - If the pattern ends with a slash, it is removed for the purpose of the
+ *   following description, but it would only find a match with a directory. In
+ *   other words, foo/ will match a directory foo and paths underneath it, but
+ *   will not match a regular file or a symbolic link foo (this is consistent
+ *   with the way how pathspec works in general in git).
+ *
+ * - If the pattern does not contain a slash /, git treats it as a shell glob
+ *   pattern and checks for a match against the pathname without leading
+ *   directories.
+ *
+ * - Otherwise, git treats the pattern as a shell glob suitable for consumption
+ *   by fnmatch(3) with the FNM_PATHNAME flag: wildcards in the pattern will
+ *   not match a / in the pathname. For example, "Documentation/\*.html" matches
+ *   "Documentation/git.html" but not "Documentation/ppc/ppc.html". A leading
+ *   slash matches the beginning of the pathname; for example, "/\*.c" matches
+ *   "cat-file.c" but not "mozilla-sha1/sha1.c".
+ */
+
+/*
+ * This will return GIT_SUCCESS if the spec was filled out,
+ * GIT_ENOTFOUND if the fnmatch does not require matching, or
+ * another error code there was an actual problem.
+ */
+static int parse_fnmatch(
+	git_attr_fnmatch *spec,
+	const char **base)
+{
+	const char *pattern;
+	const char *scan;
+	int slash_count;
+	int error = GIT_SUCCESS;
+
+	assert(base && *base);
+
+	pattern = *base;
+
+	while (isspace(*pattern)) pattern++;
+	if (!*pattern || *pattern == '#') {
+		error = GIT_ENOTFOUND;
+		goto skip_to_eol;
+	}
+
+	if (*pattern == '!') {
+		spec->negative = 1;
+		pattern++;
+	} else {
+		spec->negative = 0;
+	}
+
+	spec->fullpath = 0;
+	slash_count = 0;
+	for (scan = pattern; *scan != '\0'; ++scan) {
+		if (isspace(*scan) && *(scan - 1) != '\\')
+			break;
+
+		if (*scan == '/') {
+			spec->fullpath = 1;
+			slash_count++;
+		}
+	}
+
+	*base = scan;
+	spec->length = scan - pattern;
+	spec->pattern = git__strndup(pattern, spec->length);
+
+	if (!spec->pattern) {
+		error = GIT_ENOMEM;
+		goto skip_to_eol;
+	} else {
+		char *from = spec->pattern, *to = spec->pattern;
+		while (*from) {
+			if (*from == '\\') {
+				from++;
+				spec->length--;
+			}
+			*to++ = *from++;
+		}
+		*to = '\0';
+	}
+
+	if (pattern[spec->length - 1] == '/') {
+		spec->length--;
+		spec->pattern[spec->length] = '\0';
+		spec->directory = 1;
+		if (--slash_count <= 0)
+			spec->fullpath = 0;
+	} else {
+		spec->directory = 0;
+	}
+
+	return GIT_SUCCESS;
+
+skip_to_eol:
+	/* skip to end of line */
+	while (*pattern && *pattern != '\n') pattern++;
+	if (*pattern == '\n') pattern++;
+	*base = pattern;
+
+	return error;
+}
+
+static int sort_by_hash_and_name(const void *a_raw, const void *b_raw)
+{
+	const git_attr_name *a = a_raw;
+	const git_attr_name *b = b_raw;
+
+	if (b->name_hash < a->name_hash)
+		return 1;
+	else if (b->name_hash > a->name_hash)
+		return -1;
+	else
+		return strcmp(b->name, a->name);
+}
+
+static int parse_assigns(
+	git_vector *assigns,
+	const char **base)
+{
+	int error = GIT_SUCCESS;
+	const char *scan = *base;
+	git_attr_assignment *assign = NULL;
+
+	assert(assigns && !assigns->length);
+
+	while (*scan && *scan != '\n') {
+		const char *name_start, *value_start;
+
+		/* skip leading blanks */
+		while (isspace(*scan) && *scan != '\n') scan++;
+
+		/* allocate assign if needed */
+		if (!assign) {
+			assign = git__calloc(1, sizeof(git_attr_assignment));
+			if (!assign) {
+				error = GIT_ENOMEM;
+				break;
+			}
+		}
+
+		assign->name_hash = 5381;
+		assign->value = GIT_ATTR_TRUE;
+		assign->is_allocated = 0;
+
+		/* look for magic name prefixes */
+		if (*scan == '-') {
+			assign->value = GIT_ATTR_FALSE;
+			scan++;
+		} else if (*scan == '!') {
+			assign->value = NULL; /* explicit unspecified state */
+			scan++;
+		} else if (*scan == '#') /* comment rest of line */
+			break;
+
+		/* find the name */
+		name_start = scan;
+		while (*scan && !isspace(*scan) && *scan != '=') {
+			assign->name_hash =
+				((assign->name_hash << 5) + assign->name_hash) + *scan;
+			scan++;
+		}
+		assign->name_len = scan - name_start;
+		if (assign->name_len <= 0) {
+			/* must have found lone prefix (" - ") or leading = ("=foo")
+			 * or end of buffer -- advance until whitespace and continue
+			 */
+			while (*scan && !isspace(*scan)) scan++;
+			continue;
+		}
+
+		/* if there is an equals sign, find the value */
+		if (*scan == '=') {
+			for (value_start = ++scan; *scan && !isspace(*scan); ++scan);
+
+			/* if we found a value, allocate permanent storage for it */
+			if (scan > value_start) {
+				assign->value = git__strndup(value_start, scan - value_start);
+				if (!assign->value) {
+					error = GIT_ENOMEM;
+					break;
+				} else {
+					assign->is_allocated = 1;
+				}
+			}
+		}
+
+		/* allocate permanent storage for name */
+		assign->name = git__strndup(name_start, assign->name_len);
+		if (!assign->name) {
+			error = GIT_ENOMEM;
+			break;
+		}
+
+		/* insert allocated assign into vector */
+		error = git_vector_insert(assigns, assign);
+		if (error < GIT_SUCCESS)
+			break;
+
+		/* clear assign since it is now "owned" by the vector */
+		assign = NULL;
+	}
+
+	if (!assigns->length)
+		error = git__throw(GIT_ENOTFOUND, "No attribute assignments found for rule");
+	else {
+		assigns->_cmp = sort_by_hash_and_name;
+		git_vector_sort(assigns);
+	}
+
+	if (assign != NULL) {
+		git__free(assign->name);
+		if (assign->is_allocated)
+			git__free((void *)assign->value);
+		git__free(assign);
+	}
+
+	while (*scan && *scan != '\n') scan++;
+	*base = scan;
+
+	return error;
+}
+
+static int free_rule(git_attr_rule *rule)
+{
+	unsigned int i;
+	git_attr_assignment *assign;
+
+	if (!rule)
+		return GIT_SUCCESS;
+
+	git__free(rule->match.pattern);
+	rule->match.pattern = NULL;
+	rule->match.length = 0;
+
+	git_vector_foreach(&rule->assigns, i, assign) {
+		git__free(assign->name);
+		assign->name = NULL;
+
+		if (assign->is_allocated) {
+			git__free((void *)assign->value);
+			assign->value = NULL;
+		}
+	}
+
+	return GIT_SUCCESS;
+}
diff --git a/src/attr_file.h b/src/attr_file.h
new file mode 100644
index 000000000..4774f148c
--- /dev/null
+++ b/src/attr_file.h
@@ -0,0 +1,87 @@
+/*
+ * Copyright (C) 2009-2011 the libgit2 contributors
+ *
+ * This file is part of libgit2, distributed under the GNU GPL v2 with
+ * a Linking Exception. For full terms see the included COPYING file.
+ */
+#ifndef INCLUDE_attr_file_h__
+#define INCLUDE_attr_file_h__
+
+#include "git2/attr.h"
+#include "vector.h"
+
+typedef struct {
+	char *pattern;
+	size_t length;
+	int negative;
+	int directory;
+	int fullpath;
+} git_attr_fnmatch;
+
+typedef struct {
+	const char *name;
+	unsigned long name_hash;
+} git_attr_name;
+
+typedef struct {
+	char *name;
+    unsigned long name_hash;
+	size_t name_len;
+    const char *value;
+	int is_allocated;
+} git_attr_assignment;
+
+typedef struct {
+	git_attr_fnmatch match;
+	git_vector assigns; /* <git_attr_assignment*> */
+} git_attr_rule;
+
+typedef struct {
+	char *path;
+	git_vector rules; /* <git_attr_rule*> */
+} git_attr_file;
+
+typedef struct {
+	const char *path;
+	const char *basename;
+	int is_dir;
+} git_attr_path;
+
+/*
+ * git_attr_file API
+ */
+
+extern int git_attr_file__from_buffer(git_attr_file **out, const char *buf);
+extern int git_attr_file__from_file(git_attr_file **out, const char *path);
+
+extern void git_attr_file__free(git_attr_file *file);
+
+extern int git_attr_file__lookup_one(
+	git_attr_file *file,
+	const git_attr_path *path,
+	const char *attr,
+	const char **value);
+
+/* loop over rules in file from bottom to top */
+#define git_attr_file__foreach_matching_rule(file, path, iter, rule)	\
+	git_vector_rforeach(&(file)->rules, (iter), (rule)) \
+		if (git_attr_rule__match_path((rule), (path)) == GIT_SUCCESS)
+
+extern unsigned long git_attr_file__name_hash(const char *name);
+
+
+/*
+ * other utilities
+ */
+
+extern int git_attr_rule__match_path(
+	git_attr_rule *rule,
+	const git_attr_path *path);
+
+extern git_attr_assignment *git_attr_rule__lookup_assignment(
+	git_attr_rule *rule, const char *name);
+
+extern int git_attr_path__init(
+	git_attr_path *info, const char *path);
+
+#endif
diff --git a/src/hashtable.c b/src/hashtable.c
index 15d173992..f836f166d 100644
--- a/src/hashtable.c
+++ b/src/hashtable.c
@@ -241,3 +241,17 @@ int git_hashtable_merge(git_hashtable *self, git_hashtable *other)
 	return insert_nodes(self, other->nodes, other->key_count);
 }
 
+
+/**
+ * Standard string
+ */
+uint32_t git_hash__strhash_cb(const void *key, int hash_id)
+{
+	static uint32_t hash_seeds[GIT_HASHTABLE_HASHES] = {
+		2147483647,
+		0x5d20bb23,
+		0x7daaab3c
+	};
+
+	return git__hash(key, strlen((const char *)key), hash_seeds[hash_id]);
+}
diff --git a/src/hashtable.h b/src/hashtable.h
index f0ca3ebd2..485b17aa6 100644
--- a/src/hashtable.h
+++ b/src/hashtable.h
@@ -76,5 +76,12 @@ GIT_INLINE(int) git_hashtable_insert(git_hashtable *h, const void *key, void *va
 	_node->key = NULL; _node->value = NULL; _self->key_count--;\
 }
 
+/*
+ * If you want a hashtable with standard string keys, you can
+ * just pass git_hash__strcmp_cb and git_hash__strhash_cb to
+ * git_hashtable_alloc.
+ */
+#define git_hash__strcmp_cb git__strcmp_cb
+extern uint32_t git_hash__strhash_cb(const void *key, int hash_id);
 
 #endif
diff --git a/src/refs.c b/src/refs.c
index 8c3f700ad..cf76b23be 100644
--- a/src/refs.c
+++ b/src/refs.c
@@ -31,17 +31,6 @@ struct packref {
 
 static const int default_table_size = 32;
 
-static uint32_t reftable_hash(const void *key, int hash_id)
-{
-	static uint32_t hash_seeds[GIT_HASHTABLE_HASHES] = {
-		2147483647,
-		0x5d20bb23,
-		0x7daaab3c
-	};
-
-	return git__hash(key, strlen((const char *)key), hash_seeds[hash_id]);
-}
-
 static int reference_read(
 	git_fbuffer *file_content,
 	time_t *mtime,
@@ -443,9 +432,7 @@ static int packed_load(git_repository *repo)
 	/* First we make sure we have allocated the hash table */
 	if (ref_cache->packfile == NULL) {
 		ref_cache->packfile = git_hashtable_alloc(
-			default_table_size,
-			reftable_hash,
-			(git_hash_keyeq_ptr)&git__strcmp_cb);
+			default_table_size, git_hash__strhash_cb, git_hash__strcmp_cb);
 
 		if (ref_cache->packfile == NULL) {
 			error = GIT_ENOMEM;
diff --git a/src/repository.c b/src/repository.c
index 67afa2ee2..e0d4c6387 100644
--- a/src/repository.c
+++ b/src/repository.c
@@ -59,6 +59,7 @@ void git_repository_free(git_repository *repo)
 
 	git_cache_free(&repo->objects);
 	git_repository__refcache_free(&repo->references);
+	git_repository__attr_cache_free(&repo->attrcache);
 
 	git__free(repo->path_repository);
 	git__free(repo->workdir);
diff --git a/src/repository.h b/src/repository.h
index c3a9a5c60..5274fc1d0 100644
--- a/src/repository.h
+++ b/src/repository.h
@@ -19,6 +19,7 @@
 #include "refs.h"
 #include "buffer.h"
 #include "odb.h"
+#include "attr.h"
 
 #define DOT_GIT ".git"
 #define GIT_DIR DOT_GIT "/"
@@ -38,6 +39,7 @@ struct git_repository {
 
 	git_cache objects;
 	git_refcache references;
+	git_attr_cache attrcache;
 
 	char *path_repository;
 	char *workdir;
diff --git a/src/util.h b/src/util.h
index 4b1104b7b..be978a6a5 100644
--- a/src/util.h
+++ b/src/util.h
@@ -109,6 +109,7 @@ extern void **git__bsearch(const void *key, void **base, size_t nmemb,
 	int (*compar)(const void *, const void *));
 
 extern int git__strcmp_cb(const void *a, const void *b);
+extern uint32_t git__strhash_cb(const void *key, int hash_id);
 
 typedef struct {
 	short refcount;
diff --git a/src/vector.c b/src/vector.c
index 123aae8e6..e745d77dd 100644
--- a/src/vector.c
+++ b/src/vector.c
@@ -29,7 +29,12 @@ static int resize_vector(git_vector *v)
 void git_vector_free(git_vector *v)
 {
 	assert(v);
+
 	git__free(v->contents);
+	v->contents = NULL;
+
+	v->length = 0;
+	v->_alloc_size = 0;
 }
 
 int git_vector_init(git_vector *v, unsigned int initial_size, git_vector_cmp cmp)
diff --git a/src/vector.h b/src/vector.h
index 08f5a501c..4c053e6ae 100644
--- a/src/vector.h
+++ b/src/vector.h
@@ -19,6 +19,8 @@ typedef struct git_vector {
 	int sorted;
 } git_vector;
 
+#define GIT_VECTOR_INIT {0}
+
 int git_vector_init(git_vector *v, unsigned int initial_size, git_vector_cmp cmp);
 void git_vector_free(git_vector *v);
 void git_vector_clear(git_vector *v);
@@ -39,6 +41,9 @@ GIT_INLINE(void *) git_vector_get(git_vector *v, unsigned int position)
 #define git_vector_foreach(v, iter, elem)	\
 	for ((iter) = 0; (iter) < (v)->length && ((elem) = (v)->contents[(iter)], 1); (iter)++ )
 
+#define git_vector_rforeach(v, iter, elem)	\
+	for ((iter) = (v)->length; (iter) > 0 && ((elem) = (v)->contents[(iter)-1], 1); (iter)-- )
+
 int git_vector_insert(git_vector *v, void *element);
 int git_vector_remove(git_vector *v, unsigned int idx);
 void git_vector_uniq(git_vector *v);