Re: [PATCH] cat-file: speed up default format
- From
Jeff King <peff@peff.net>
- Date
- Jun 15, 2026, 17:06 UTC
- Message-ID
- <20260615170652.GB91269@coredump.intra.peff.net>
- In-Reply-To
- <20260615165326.GA91269@coredump.intra.peff.net>
On Mon, Jun 15, 2026 at 12:53:26PM -0400, Jeff King wrote:
> It uses per-atom callback functions which is nice and clean, though we > might be able to do even better with a big ugly switch() statement.
Being the curious sort, I swapped it out for a big switch statement. Patch below, but it does not seem to be any faster.
So the bottom line is I think you could gain a little bit of performance by pre-parsing (versus strbuf_expand() on each object). Around 3% for something that actually looks at the objects, though more like 15% if for just dumping the objectnames.
IMHO that is probably not worth it for a custom parsing system just for cat-file. But if we were to finally unify ref-filter and cat-file (and even --pretty=format) then it would probably worth doing this kind of pre-parsing.
---
diff --git a/builtin/cat-file.c b/builtin/cat-file.c index 9cc7ec7a6f..da6ecc61f9 100644 --- a/builtin/cat-file.c +++ b/builtin/cat-file.c @@ -321,7 +321,17 @@ struct expand_data { #define EXPAND_DATA_INIT { .mode = S_IFINVALID } struct format_item { - void (*add)(struct format_item *item, struct strbuf *sb, struct expand_data *data); + enum { + FORMAT_TYPE_END = 0, + FORMAT_TYPE_LITERAL, + FORMAT_TYPE_OBJECTNAME, + FORMAT_TYPE_OBJECTTYPE, + FORMAT_TYPE_OBJECTSIZE, + FORMAT_TYPE_OBJECTSIZE_DISK, + FORMAT_TYPE_REST, + FORMAT_TYPE_DELTABASE, + FORMAT_TYPE_OBJECTMODE, + } type; union { struct { const char *p; @@ -336,55 +346,6 @@ struct format_item { */ }; -static void objectname_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - strbuf_add_oid_hex(sb, &data->oid); -} - -static void objecttype_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - strbuf_addstr(sb, type_name(data->type)); -} - -static void objectsize_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - strbuf_add_uint(sb, data->size); -} - -static void objectsize_disk_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - strbuf_add_uint(sb, data->disk_size); -} - -static void rest_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - strbuf_addstr(sb, data->rest); -} - -static void deltabase_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - strbuf_add_oid_hex(sb, &data->delta_base_oid); -} - -static void objectmode_add(struct format_item *item UNUSED, - struct strbuf *sb, struct expand_data *data) -{ - if (data->mode != S_IFINVALID) - strbuf_addf(sb, "%06o", data->mode); -} - -static void literal_add(struct format_item *item, - struct strbuf *sb, struct expand_data *data UNUSED) -{ - strbuf_add(sb, item->u.literal.p, item->u.literal.len); -} - static int is_atom(const char *atom, const char *s, int slen) { int alen = strlen(atom); @@ -395,24 +356,24 @@ static int parse_atom(struct format_item *fmt, const char *atom, int len, struct expand_data *data) { if (is_atom("objectname", atom, len)) { - fmt->add = objectname_add; + fmt->type = FORMAT_TYPE_OBJECTNAME; } else if (is_atom("objecttype", atom, len)) { data->info.typep = &data->type; - fmt->add = objecttype_add; + fmt->type = FORMAT_TYPE_OBJECTTYPE; } else if (is_atom("objectsize", atom, len)) { data->info.sizep = &data->size; - fmt->add = objectsize_add; + fmt->type = FORMAT_TYPE_OBJECTSIZE; } else if (is_atom("objectsize:disk", atom, len)) { data->info.disk_sizep = &data->disk_size; - fmt->add = objectsize_disk_add; + fmt->type = FORMAT_TYPE_OBJECTSIZE_DISK; } else if (is_atom("rest", atom, len)) { data->split_on_whitespace = 1; - fmt->add = rest_add; + fmt->type = FORMAT_TYPE_REST; } else if (is_atom("deltabase", atom, len)) { data->info.delta_base_oid = &data->delta_base_oid; - fmt->add = deltabase_add; + fmt->type = FORMAT_TYPE_DELTABASE; } else if (is_atom("objectmode", atom, len)) { - fmt->add = objectmode_add; + fmt->type = FORMAT_TYPE_OBJECTMODE; } else return 0; return 1; @@ -430,7 +391,7 @@ static struct format_item *parse_format(const char *start, if (percent != start) { ALLOC_GROW(ret, nr + 1, alloc); - ret[nr].add = literal_add; + ret[nr].type = FORMAT_TYPE_LITERAL; ret[nr].u.literal.p = start; ret[nr].u.literal.len = percent - start; nr++; @@ -443,7 +404,7 @@ static struct format_item *parse_format(const char *start, ALLOC_GROW(ret, nr + 1, alloc); if (skip_prefix(start, "%", &start) || *start != '(') { - ret[nr].add = literal_add; + ret[nr].type = FORMAT_TYPE_LITERAL; ret[nr].u.literal.p = "%"; ret[nr].u.literal.len = 1; } else if ((end = strchr(start + 1, ')')) && @@ -456,16 +417,44 @@ static struct format_item *parse_format(const char *start, } ALLOC_GROW(ret, nr + 1, alloc); - ret[nr].add = NULL; + ret[nr].type = FORMAT_TYPE_END; return ret; } static void expand_format(struct strbuf *sb, struct format_item *fmt, struct expand_data *data) { - for (; fmt->add; fmt++) - fmt->add(fmt, sb, data); + for (; fmt->type; fmt++) + switch (fmt->type) { + case FORMAT_TYPE_END: + BUG("we should have already left the loop!"); + break; + case FORMAT_TYPE_OBJECTNAME: + strbuf_add_oid_hex(sb, &data->oid); + break; + case FORMAT_TYPE_OBJECTTYPE: + strbuf_addstr(sb, type_name(data->type)); + break; + case FORMAT_TYPE_OBJECTSIZE: + strbuf_add_uint(sb, data->size); + break; + case FORMAT_TYPE_OBJECTSIZE_DISK: + strbuf_add_uint(sb, data->disk_size); + break; + case FORMAT_TYPE_REST: + strbuf_addstr(sb, data->rest); + break; + case FORMAT_TYPE_DELTABASE: + strbuf_add_oid_hex(sb, &data->delta_base_oid); + break; + case FORMAT_TYPE_OBJECTMODE: + if (data->mode != S_IFINVALID) + strbuf_addf(sb, "%06o", data->mode); + break; + case FORMAT_TYPE_LITERAL: + strbuf_add(sb, fmt->u.literal.p, fmt->u.literal.len); + } } static void batch_write(struct batch_options *opt, const void *data, int len)