trace/mercury_trace_cmd_browsing.c:
Change the dump command to allow the name of the variable
whose value is to be dumped to be followed by a path,
specifying the *part* of the variable's value to be printed.
trace/mercury_trace_vars.[ch]:
To make that possible, provide a version of an existing function
that returns an extra value that the new code in
mercury_trace_cmd_browsing.c needs, and another extra value
that conceptually belongs with it.
Mark strange code with an XXX.
NEWS.md:
Announce both this new capability and a just-now contributes predicate.
library/pqueue.m:
Make the just-now contributed code more readable.
tests/debugger/browser_test.{m,inp,exp3}:
Extend this test case to test the new functionality.
The usual mdb "dump" command puts every function symbol on its own line.
This guarantees that we generate any line that is too long to be displayed
on terminals, but it also generates output that is too stretched out
vertically for its structure to be readily apparent. Dumping the term
as a doc allows pretty_printer.m to put as many function symbols on a line
as would fit, without exceeding the maximum line length.
browser/browse.m:
Add save_term_to_file_doc, a way to save a (possibly synthetic)
browser term in a file, using an interface that works the same way
the predicate that saves browser terms as XML.
Inline a predicate at its only call site. Improve variable names.
trace/mercury_trace_browse.[ch]:
Add MR_trace_save_term_as_doc, as an interface function between
save_term_to_file_doc and the code of mercury_trace_cmd_browsing.c.
trace/mercury_trace_cmd_browsing.c:
Add support for a new -p/--prettyprint flag to the mdb "dump" command,
which asks for the given term to be dumped as a pretty_printer doc.
doc/user_guide.texi:
NEWS:
Document the new option.
library/pretty_printer.m:
NEWS:
Rename write_as_doc to write_doc_formatted, and fix its argument type.
tests/debugger/browser_test.inp:
Dump a term that we already dumped with "dump" with "dump -x" and
"dump -p" as well.
tests/debugger/browser_test.m:
Put the code to remove the files we are going to dump to
and then later to print the files we have dumped to into separate
predicates. This keeps most (but not all) line numbers unchanged
even though we now dump to more files.
tests/debugger/browser_test.exp3:
Update this file to account both for the extra output from the just-added
dump commands, and for the changes in line numbers.
The 'browse --xml' command has not worked with current versions of
xsltproc for quite some time, but we have not received any bug reports,
nor has anyone tried to fix it. We have a method for interactively
exploring a term in 'browse --web' so IMHO there is no need to keep
support for 'browse --xml'.
browser/browse.m:
browser/browser_info.m:
browser/declarative_user.m:
trace/mercury_trace_browse.c:
trace/mercury_trace_browse.h:
trace/mercury_trace_cmd_browsing.c:
trace/mercury_trace_cmd_parameter.c:
trace/mercury_trace_cmd_parameter.h:
trace/mercury_trace_internal.c:
Delete code.
doc/mdb_categories:
doc/user_guide.texi:
Delete documentation.
configure.ac:
Don't search for a XUL browser and xsltproc.
scripts/mdbrc.in:
Delete 'xml_browser_cmd' and 'xml_tmp_filename' lines.
scripts/xul_tree.xsl:
Delete now unused file.
scripts/Mmakefile:
Conform to deletions.
tests/debugger/Mmakefile:
tests/debugger/browser_test.exp:
tests/debugger/browser_test.exp3:
tests/debugger/browser_test.inp:
tests/debugger/mdb_command_test.inp:
tests/debugger/save.exp2:
tests/declarative_debugger/browse_arg.exp:
tests/declarative_debugger/browse_arg.inp:
Don't test 'browse --xml' any longer.
extras/xml_stylesheets/README:
Delete reference to 'browse --xml' command.
NEWS:
Announce change.
LICENSE:
Mention files derived from musl library.
runtime/process_getopt:
Process files from ../getopt/*.[ch] to produce
mercury_getopt.h, mercury_getopt.c, mercury_getopt_long.c.
runtime/mercury_getopt.h:
runtime/mercury_getopt.c:
runtime/mercury_getopt_long.c:
Add files generated by process_getopt.
runtime/Mmakefile:
Update list of source files.
util/Mmakefile:
Use musl getopt if the system C library does not have getopt().
tests/debugger/browser_test.exp:
tests/debugger/browser_test.exp2:
tests/debugger/browser_test.exp3:
tests/hard_coded/runtime_opt.exp:
Update expected error messages to match new getopt() output.
changed exp output
tests/accumulator/*.m:
tests/analysis_*/*.m:
tests/benchmarks*/*.m:
tests/debugger*/*.{m,exp,inp}:
tests/declarative_debugger*/*.{m,exp,inp}:
tests/dppd*/*.m:
tests/exceptions*/*.m:
tests/general*/*.m:
tests/grade_subdirs*/*.m:
tests/hard_coded*/*.m:
Make these tests use four-space indentation, and ensure that
each module is imported on its own line. (I intend to use the latter
to figure out which subdirectories' tests can be executed in parallel.)
These changes usually move code to different lines. For the debugger tests,
specify the new line numbers in .inp files and expect them in .exp files.
Estimated hours taken: 2
Branches: main
Make the mdb command "dump" give you a message confirming what it did.
trace/mercur_trace_cmd_browsing.c:
Make the mdb command "dump" give you a message confirming what it did,
unless the user gives the new -q or --quiet option.
doc/user_guide.texi:
Document the new option.
scripts/mdb_grep:
scripts/mdb_open:
Specify -q to keep old behavior.
tests/debugger/browser_test.{inp,exp,exp3}:
Update this test case to test the dump command both with and
without -q.
Estimated hours taken: 3
Branches: main
When the user asks for the printing of a non-existent part of a term, make
mdb print the path up to the nonexistent part as part of the error message.
trace/mercury_trace_var.[ch]:
Give to the function that generates those error messages a pointer
not just to the bad part of the path, but to the full path as well,
so it can print the good part.
Remove fixed limits on the size of error messages.
trace/mercury_trace_spy.c:
Conform to the change in mercury_trace_var.h.
runtime/mercury_memory.[ch]:
Provide a mechanism for dynamically resizing character buffers
for use by mercury_trace_vars.c. We use this instead of snprint
because snprint isn't available or doesn't work properly on some
platforms.
Document a previously overlooked function.
tests/debugger/browser_test.exp3:
tests/debugger/field_names.exp:
Update these test cases to expect the new, more informative
error messages.
Estimated hours taken: 1
Branches: main
library/pretty_printer.m:
The depth limit was being prematurely decremented when formatting
terms, causing one too few functors to be displayed before
truncation. This bug caused everything to be printed as "..."
at depth 1, rather than just showing the top-level functor.
tests/debugger/browse_pretty.exp:
tests/debugger/browser_test.exp3:
tests/debugger/declarative/sort.exp:
tests/hard_coded/test_pretty_printer.exp:
tests/hard_coded/test_pretty_printer_defaults.exp:
Update expected output files.
tests/hard_coded/test_pretty_printer.m:
Test the depth=1 case.
Estimated hours taken: 5 (mostly testing)
Branches: main
browser/browse.m:
Try to handle pretty-printing of functor names containing '.'s
correctly (the default pretty-printer formatting escapes names
containing '.'s).
library/list.m:
Improve the pretty-printer formatting of lists by preventing
the output of ellipsis when all items in the list have been
printed.
library/pretty_printer.m:
Special-case printing of "." as an infix operator by not
surrounding it by spaces.
tests/debugger/browse_pretty.exp:
tests/debugger/browser_test.exp:
tests/debugger/declarative/change_search.exp:
Updated to reflect output of new pretty printer.
tests/debugger/browser_test.exp3:
Added a new possible output (for new pretty printer).
