[pve-devel] [PATCH pve-zsync 1/2] rename encode_cron to parse_cron

[pve-devel] [PATCH pve-zsync 1/2] rename encode_cron to parse_cron

[Fabian Ebner]

The old name is confusing, because "encode" is usually not used when
converting from text to a data structure.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
---
 pve-zsync | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/pve-zsync b/pve-zsync
index b466076..7246336 100755
--- a/pve-zsync
+++ b/pve-zsync
@@ -232,7 +232,7 @@ sub read_cron {
 
     my $text = read_file($CRONJOBS, 0);
 
-    return encode_cron(@{$text});
+    return parse_cron(@{$text});
 }
 
 sub parse_argv {
@@ -302,7 +302,7 @@ sub add_state_to_job {
     return $job;
 }
 
-sub encode_cron {
+sub parse_cron {
     my (@text) = @_;
 
     my $cfg = {};
-- 
2.30.2

[pve-devel] [PATCH pve-zsync 2/2] parse cron: handle additional whitespace

[Fabian Ebner]

Can only happen by manually editing AFAICT, but cron does execute the
jobs, so just be a bit less restrictive when parsing.

Reported in the community forum:
https://forum.proxmox.com/threads/105254/

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
---
 pve-zsync | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/pve-zsync b/pve-zsync
index 7246336..f69e126 100755
--- a/pve-zsync
+++ b/pve-zsync
@@ -308,8 +308,7 @@ sub parse_cron {
     my $cfg = {};
 
     while (my $line = shift(@text)) {
-
-	my @arg = split('\s', $line);
+	my @arg = split('\s+', $line);
 	my $param = parse_argv(@arg);
 
 	if ($param->{source} && $param->{dest}) {
-- 
2.30.2

Re: [pve-devel] [PATCH pve-zsync 2/2] parse cron: handle additional whitespace

[Thomas Lamprecht]

On 21.02.22 10:07, Fabian Ebner wrote:
> Can only happen by manually editing AFAICT, but cron does execute the
> jobs, so just be a bit less restrictive when parsing.

I mean, it won't get more broken as already but can we avoid white-space splitting
on command arguments in general?

We could use the same underlying helper as PVE::Tools::split_args uses,
Text::ParseWords::shellwords (included in perl-modules directly)..

> 
> Reported in the community forum:
> https://forum.proxmox.com/threads/105254/
> 
> Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
> ---
>  pve-zsync | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/pve-zsync b/pve-zsync
> index 7246336..f69e126 100755
> --- a/pve-zsync
> +++ b/pve-zsync
> @@ -308,8 +308,7 @@ sub parse_cron {
>      my $cfg = {};
>  
>      while (my $line = shift(@text)) {
> -
> -	my @arg = split('\s', $line);
> +	my @arg = split('\s+', $line);
>  	my $param = parse_argv(@arg);
>  
>  	if ($param->{source} && $param->{dest}) {

[pve-devel] applied: [PATCH pve-zsync 1/2] rename encode_cron to parse_cron

[Thomas Lamprecht]

On 21.02.22 10:07, Fabian Ebner wrote:
> The old name is confusing, because "encode" is usually not used when
> converting from text to a data structure.
> 
> Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
> ---
>  pve-zsync | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
>

applied, thanks!