Add/clear rules dynamically about workflow HOT 4 CLOSED

Hello Jonathan,

some things you can do, some will require work.

If you edit workflow.config, the new config will take effect immediately. You can already use # to comment rules. You can enable them at runtime by un-commenting them. The only thing you cannot do is have it "forget" that a rule was processed and have it processed again. Yet what you can do is "touch" all the files and the rule will run again.

One thing you cannot do at runtime is call workflow.py -c . That is because workflow.cache is a shelve object and gets corrupted is accessed concurrently.

Is this sufficient? Give it a try.

Massimo

On May 31, 2012, at 8:10 PM, Jonathan Paugh wrote:

Great stuff! This is a handy little script!
I'd like to see a way to add/clear rules from a running instance of workflow. For example, I often want to recompile a source file I'm working on whenever I save it to disk. If I set up workflow to monitor my home directory, then use judicious globs, I could use a rule like this:

recompile: proj/tex/test-proj/*.tex: pdftex $0

But, I'd like to be able to add this to a running instance of workflow, rather than editing the config file and restarting workflow every time I want to add a new rule. Especially since I don't want it to always use this rule: just while I'm working on test-proj.

I think this could be done in 3 steps:

• watch the cache file for changes, and reload the data dict if it changes.
• add a --new option that adds a new rule, like so: ./workflow.py --new "recompile: proj/tex/test-proj/*.tex: pdftex \$0"
  This would mirror the behavior of --clear
• add an option to change the currently running workflow, rather than starting a new one.
  This would cause either --new or --clear to update the cache, but immediately shut down, rather than the norm.

I'm going to look into this and see how doable it is. (I might have a patch in a few days.) In particular, I wonder if shelve is thread-safe like that. If not, an alternate solution would be to contact the current daemon and let it update the cache itself.

---

Reply to this email directly or view it on GitHub:
#4

from workflow.

Alright, that sounds okay. If I can figure out how, I'd still like to do it from the command line, though. If I make any progress, I'll let ya know. (I haven't worked on a LaTeX project for a while--but next time, I'll have workflow. :-)

from workflow.

I implemented your suggestion:

start the daemon
$ python workflow.py -d
clear a run at runtime
$ python workflow.py -c myrule

this creates the file .workflow.myrule.clear, the daemon reads it, clears workflow.cache and deleted the tmp file.
I like your idea because you can have a rule that "forgets" another rule by creating the tmp file.

Please check it.

Massimo

On May 31, 2012, at 8:27 PM, Jonathan Paugh wrote:

Update: I see that shelve can't stand having multiple writers to a single cache. A workaround would be to generate temporary config file(s) in the folder specified by --f with names like .add-workflow.PID or clear-workflow.PID where PID is the process id of the adding/clearing process. Idk. I guess it's a bit of a hack however it's done.

---

Reply to this email directly or view it on GitHub:
#4 (comment)

from workflow.

I tested this out, and I like it a lot. I used that as a starting point to add an --add option. (Check my pull request).

from workflow.