Merge pull request #187 from dbosk/send-copy-of-final-timesheet-to-TA

Improves `timesheets show`, adapts to HR+ 8

src/nytid/cli/hr.nw

@@ -1658,7 +1658,8 @@ We let this command just print a summary of each time sheet.
 def cli_show_timesheets(<<argument [[user_regex]] to match TAs>> = ".*",
                         <<option [[reverse]] to adjust sort order>>,
                         <<option [[all]] to show all timesheets>>,
-                        <<option for CSV delimiter>>):
+                        <<option for CSV delimiter>>,
+                        <<option for [[date_regex]] to filter by date>>,):
   """
   Shows stored time sheets for TAs.
   """
@@ -1668,6 +1669,7 @@ def cli_show_timesheets(<<argument [[user_regex]] to match TAs>> = ".*",
   <<set [[timesheets]] depending on [[all]]>>
   for timesheet in timesheets:
     <<skip if [[user_regex]] doesn't match [[timesheet]]>>
+    <<skip if [[date_regex]] doesn't match [[timesheet]]>>
     <<skip if we've already printed a timesheet for [[user]]>>
     <<print [[timesheet]]>>
 <<option [[reverse]] to adjust sort order>>=
@@ -1694,6 +1696,8 @@ timesheets = sorted(timesheets_dir.glob(f"*.json"),
 @
 
 We can check if the time sheet file matches the user's username.
+For example, the file name might be
+[[dbosk.2024-09-26T14:28:39.434834.json]].
 <<skip if [[user_regex]] doesn't match [[timesheet]]>>=
 username = timesheet.name.split(".")[0]
 if not user_pattern.match(username):
@@ -1702,14 +1706,32 @@ if not user_pattern.match(username):
 user_pattern = re.compile(user_regex)
 @
 
+In the same fashion, we want to filter on dates.
+<<skip if [[date_regex]] doesn't match [[timesheet]]>>=
+if date_regex and not date_pattern.match(timesheet.name.split(".")[1]):
+  continue
+<<timesheets iteration variables>>=
+if date_regex:
+  date_pattern = re.compile(date_regex)
+<<option for [[date_regex]] to filter by date>>=
+date_regex: Annotated[str, date_regex_opt] = None
+<<argument and option definitions>>=
+date_regex_opt = typer.Option(help="Filter time sheets by date. This is a "
+                                   "regex matching the date of the time "
+                                   "sheet.")
+@
+
+
 In the same fashion, we can check if we've already printed a time sheet for the 
 user.
 If we haven't printed one, we add the user to the set of printed users and 
 proceed to the printing step.
 We need to do like this since we might want to print the newest time sheet of 
 several matching users.
+Consequently, we don't want to skip when we want to print all time sheets for 
+the TAs.
 <<skip if we've already printed a timesheet for [[user]]>>=
-if username in printed_users:
+if username in printed_users and not all:
   continue
 else:
   printed_users.add(username)

src/nytid/cli/init.nw

@@ -411,7 +411,7 @@ The same if the UG group is empty (since that would regex match any course in
 UG).
 <<look up [[ug_path]], or skip if none exists>>=
 ug_path=$(nytid courses config ${course} ug.assistants 2>/dev/null \
-            | sed "s/.*= //")
+          | sed "s/.*= //")
 test "$?" -eq 0 || continue
 test -n "${ug_path}" || continue
 @
@@ -878,7 +878,16 @@ And if so, we can create the contract with the correct start date.
 
 \subsection{Generating time sheets for hourly TAs}
 
-The TAs that are paid by the hour must submit time sheets according to HR.
+Now with the new HR system, HR+ 8, the TAs should report their hours directly 
+in that system.
+Then the course responsible should approve them.
+This means that we no longer need the [[nytid hr timesheets generate]] 
+functionality as it was originally intended.
+But we can describe how it works, in case we can use it for something else.
+
+The idea of the previous system was as follows.
+The TAs that are paid by the hour previously had to submit time sheets 
+according to HR.
 However, then we must check them.
 It's better that we generate them automatically from the sign-up sheet, then we 
 know they're correct as long as the sign-up sheet is correct.
@@ -917,7 +926,7 @@ end date.
 Another thing to note is that we used the [[--draft]] option.
 This will generate a draft time sheet directly in the terminal.
 Without the draft option, we would get a file with the signed time sheet, you 
-can see it in \cref{FigTimesheet}.
+can see the signed time sheet in \cref{FigTimesheet}.
 
 \begin{figure}
   \centering
@@ -955,7 +964,8 @@ This way we can do calculations on the data, which we need to do (see
 
 With this construction, we can automate the generation of time sheets on a 
 monthly basis.
-<<nytid.monthly.sh>>=
+However, we no longer need this, this is just for reference.
+<<sending signed time sheets>>=
 <<timesheets variables>>
 
 for course in ${TIMESHEETS_COURSES}
@@ -1089,18 +1099,21 @@ reference.
 
 \subsection{Send time-sheet drafts to the TAs}
 
-We shouldn't send the time sheets to HR before the TAs have had a chance to 
-check them.
-They will sometimes have things to update due to changes, even though all 
+One thing we can continue doing, despite the new HR+ 8 system, is to send 
+drafts of the time sheets to the TAs.
+Then the TAs can use them as a base to report their hours in HR+ 8.
+
+The TAs will sometimes have things to update due to changes, even though all 
 changes should be made in advance.
+So we want them to update the sheet before they submit to HR+.
 The same goes for the amanuensis, we want to send them a draft of the mertid 
 that they've done.
 We want them to keep track of how many hours too many or too few they've 
 worked.
 And give them the option to report the hours to get paid.
 
 The structure of the code is similar to that of
-[[<<nytid.monthly.sh>>]].
+[[<<sending signed time sheets>>]].
 <<nytid.premonthly.sh>>=
 <<timesheet draft variables>>
 <<timesheet draft functions>>
@@ -1175,23 +1188,23 @@ Du har ${sum_hours} timmar i bokningsarket för ${course}.
 
 $(if <<[[sum_hours]] > 0>>; then
     echo "
-Den första nästa månad (${next_month}) kommer alla tider i
-bokningsarket att tidrapporteras automatiskt åt dig. För
-tillfället är det tiderna nedan (även bifogat som CSV-fil).
+Du kan rapportera in timmarna nedan i KTH HR. Gör det kring
+månadsskiftet för att få betalt i slutet av ${next_month}.
+Tiderna finns även bifogat som CSV-fil.
 " | fmt
   elif <<[[sum_hours]] < 0>>; then
     echo "
 Du har just nu fått mer betalt än du har jobbat. Du behöver
-jobba in de timmar som saknas.
+jobba in de timmar som saknas. Se nedan.
 " | fmt
   else
     echo "
 Du har ändrat pass, men alla timmar går jämnt ut.
 " | fmt
   fi)
 
-Om något är fel, uppdatera bokningsarket eller svara på detta
-mail snarast!
+Om något är fel, uppdatera bokningsarket och svara på detta
+mail!
 
           Daniel
 
@@ -1238,7 +1251,7 @@ CSV-filen.)
 $(if <<[[sum_hours]] > 0>>; then
     echo "
 Du ligger på plus och kan rapportera in ${sum_hours} timmar som
-mertid i HR+. Men svara då på detta mail och meddela att du gör det.
+mertid i KTH HR. Men svara då på detta mail och meddela att du gör det.
 Bra jobbat!
 " | fmt
   elif <<[[sum_hours]] < 0>>; then