Hi Denton
On 21/03/2020 09:21, Denton Liu wrote:
In the future, we plan on externing read_oneliner(). Future users of
read_oneliner() will want the ability to output warnings in the event
that the `path` doesn't exist. Introduce the
`READ_ONELINER_WARN_NON_EXISTENCE`
This is a rather cumbersome name, I think Junio's original suggestion of
READ_ONELINER_WARN_MISSING is nicer
flag which, if active, would issue a
warning when a file doesn't exist by skipping the `!file_exists()` check
and letting `strbuf_read_file()` handle that case.
Signed-off-by: Denton Liu <liu.denton@xxxxxxxxx>
---
sequencer.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/sequencer.c b/sequencer.c
index abb2a21e9d..92e8d38290 100644
--- a/sequencer.c
+++ b/sequencer.c
@@ -420,6 +420,7 @@ static int write_message(const void *buf, size_t len, const char *filename,
}
#define READ_ONELINER_SKIP_IF_EMPTY (1 << 0)
+#define READ_ONELINER_WARN_NON_EXISTENCE (1 << 1)
/*
* Reads a file that was presumably written by a shell script, i.e. with an
@@ -436,7 +437,7 @@ static int read_oneliner(struct strbuf *buf,
int ret = 0;
struct strbuf file_buf = STRBUF_INIT;
- if (!file_exists(path))
+ if (!(flags & READ_ONELINER_WARN_NON_EXISTENCE) && !file_exists(path))
return 0;
This isn't the fault of you patch but it would be more efficient to do
the read and then check for ENOENT/ENOTDIR rather than calling
file_exists() first
Best Wishes
Phillip
if (strbuf_read_file(&file_buf, path, 0) < 0) {
