* include/abg-comparison.h (class base_diff): New.
* include/abg-ir.h (function_decl::get_first_non_implicit_parm):
New member function.
(function_type::get_first_non_implicit_parm): Likewise.
* src/abg-comparison.cc (base_has_changed): Change parm to take
class_decl::base_spec_sptr.
(class_diff::ensure_lookup_tables_populated): Compare the base
specs not just the base classes.
(class_diff::report): Adjust. Report sub-type changes in the
member functions.
(base_diff::*): Define member functions.
(compute_diff): Define an overload for base_diff_sptr.
* src/abg-hash.cc ({function_type,
method_type}:#️⃣:operator()): Do not hash the implicit parm of
member functions.
(class_decl:#️⃣:operator()): Do not hash member types.
* src/abg-ir.cc (decl_base::decl_base): Initialize the
hashing_started_ member that got moved here from class_decl.
(decl_base::get_hash): Do not set the hash if it's being set b/c
we are in a class_decl.
(decl_base::operator==) Do not compare hashes for now. Two decls
can have different hashes and compare equal; think about an
incomplete type foo, that compares equal with a complete foo.
Their hashes will be different though. So for now, just avoid
comparing these.
(compare_function_types): Avoid comparing the implicit parameter
for member functions.
({function, method}_type::get_first_non_implicit_parm): New
definition.
* tests/data/test-read-write/test20.xml: Update.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/abg-dwarf-reader.cc (build_class_type_and_add_to_ir): Do not
declare classes before defining them because both the declaration
and the definition now compare equal.
* tests/data/test-read-dwarf/test1.abi: Update test.
* tests/data/test-read-write/test20.xml: Likewise.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-ir.h (class_decl::is_struct_): New member.
(class_decl::class_decl): Take an additional is_struct member.
(class_decl::is_struct): New getter.
* src/abg-ir.cc (class_decl::class_decl): Initialize the new
is_struct_ data member.
(class_decl::get_pretty_representation): Tell struct and class
apart.
* abg-reader.cc (read_is_struct): New static function.
(build_class_decl): Capture the struct-ness of the class.
* src/abg-dwarf-reader.cc (build_class_type_and_add_to_ir):
Likewise.
* src/abg-writer.cc (write_is_struct): New static function.
(write_class_decl): Write the struct-ness of the class.
* tests/data/test-read-dwarf/test1.abi: Update test.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/abg-dwarf-reader.cc (die_is_artificial): New static
function.
(build_class_type_and_add_to_ir, build_ir_node_from_die): Use the
die_is_artificial function to test if the function DIE is
artificial and then skip it.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/abg-dwarf-reader.cc (build_ir_node_from_die): Insert
reference, pointers and qualified types at the point where we see
them, in their scope, rather than under the scope of their
underlying type. This is because their might be first declared as
member types and we don't want to loose that information.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/abg-comparison.h (compute_diff): For class_decl_sptr, do not
compare member types and member class templates. This adds a lots
of noise that is not necessarily useful for now.
* tests/data/test-bidiff/test-qual-type0-report.txt: Update tests.
* tests/data/test-bidiff/test-struct0-report.txt: Likewise.
* tests/data/test-bidiff/test-struct1-report.txt: Likewise.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
accordingly
* include/abg-comparison.h (class diff_context): New class.
(class diff::{ctxt_, reported_once_, currently_reporting_}): New
data members.
(diff::diff): Initialize the new data members above.
(diff::{context, currently_reporting, reported_once}): New
accessors.
(compute_diff, var_diff::var_diff, pointer_diff::pointer_diff)
(reference_diff::reference_diff, qualified_type_diff)
(enum_diff:enum_diff, class_diff::class_diff)
(scope_diff::scope_diff, function_decl_diff::function_decl_diff)
(type_decl_diff::type_decl_diff, typedef_diff::typedef_diff)
(translation_unit_diff::translation_unit_diff, corpus_diff::corpus_diff):
Take an additional pointer to diff_context.
* abg-comparison.cc (diff_context::{has_diff_for,
has_diff_for_types, add_diff}): New methods.
(try_to_diff, compute_diff_for_types, compute_diff_for_decls)
(represent): Take an additional pointer to
diff_context in argument. In the later function, do not re-report
a diff if it has already been reported, or if it's being reported
already.
(var_diff::var_diff, pointer_diff::pointer_diff)
(reference_diff::reference_diff)
(qualified_type_diff::qualified_type_diff, enum_diff::enum_diff)
(class_diff::class_diff, scope_diff::scope_diff)
(function_decl_diff::function_decl_diff, type_decl::type_decl)
(typedef_diff::typedef_diff)
(translation_unit_diff::translation_unit_diff)
(corpus_diff::corpus_diff): Take an additional pointer to
diff_context in argument.
({pointer_diff, qualified_type_diff,
reference_type_diff}::report): do not re-report a diff about the
underlying type if it has already been reported, or if it's being
reported already.
(enum_diff::report): Fix this to properly use the populated lookup
tables.
(compute_diff): take an additional pointer to diff_context in
argument. For the var_decl, pointer_diff reference_type_diff,
qualified_type_diff enum_diff, scope_diff, function_decl_diff,
type_decl_diff and typedef_diff overloads, do not re-build a diff
object, if one exits already. Otherwise, record the new diff
object created so that it can be re-used later.
(enum_diff::ensure_lookup_tables_populated): Fix logic to avoid
one loop.
(class_decl::priv::{deleted_member_functions_,
inserted_member_functions_, changed_member_function_}): New
members to support reporting about member functions changes.
(class_decl::{lookup_tables_empty, clear_lookup_tables, length):
Update for the new additions above.
(class_decl::ensure_lookup_tables_populated): Likewise. Fix to
properly use the lookup tables and also avoid a going through
several loops to compute the changed members.
(class_decl::report): Flip a switch to make the beginning and end
of the reporting, in the context. Also, do not try to report
again, if we were already reporting this diff. Fix quite some
spots to properly use the lookup tables.
(scope_diff::ensure_lookup_tables_populated): Skip decl-only
classes during comparison. Fix some thinkos. Fix logic to avoid a
loop.
(scope_diff::report): Adjust to pass a context to
compute_diff_for_types.
(function_decl_diff::ensure_lookup_tables_populated): Fix logic to
avoid a loop.
(function_decl_diff::report): Adjust call to
compute_diff_for_types to pass the context.
(typedef::report): Avoid re-reporting the diff of the underlying
types, if we are already reporting it.
(corpus_diff::priv::ensure_lookup_tables_populated): Use the
pretty representation of the function rather than its name to key
the maps of deleted and added functions. Fix logic to avoid going
through an additional loop for the changed functions.
(corpus_diff::report): Add a title for removed/added/changed
functions. Fix indentation for added/removed/changed functions.
* include/abg-ir.h (class_decl::comparison_started_): New member
* src/abg-dwarf-reader.cc (is_public_decl): Style fix.
(is_declaration_only_): New static function.
(build_class_type_and_add_to_ir): Create decl-only classes (IR) for
classes flagged as declaration-only in the DWARF.
* src/abg-hash.cc (class_decl:#️⃣:operator()): Do not forget to
include the "is_declaration_only" flag into the hashing.
* src/abg-ir.cc (class_decl::operator==): Look through decl-only
classes to get their definitions and compare the definitions
instead. Avoid comparing member types and fns if the comparison
of this type has already started.
* src/abg-reader.cc (build_class_decl): Set the definition of a
declaration, when we see it.
* tests/data/test-bidiff/test-qual-type0-report.txt: Update.
* tests/data/test-bidiff/test-struct0-report.txt: Likewise.
* tests/data/test-bidiff/test-struct1-report.txt: Likewise.
signed-off-by: Dodji Seketeli <dodji@redhat.com>
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-diff-utils.h (end_of_frr_d_path_in_k_plus_delta):
Favour moving left when the two abscissas at the previous steps
are equal.
(compute_diff): Update the length of the shortest edit script when
the size of one of the inputs is zero.
* tests/test-core-diff.cc (in_out_spec): Add a new input to diff
two sequences for regression testing.
* tests/data/test-core-diff/report13.txt: New reference for
the comparison of the new regression test above.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/abg-dwarf-reader.cc (insert_decl_into_ir_under_scope)
(build_namespace_decl_and_add_to_ir, build_enum_type)
(build_class_type_and_add_to_ir, build_ir_node_from_die)
(build_ir_node_from_die): Remove.
* tests/data/test-read-dwarf/test0.abi: Update because now type
IDs can be used before they are defined.
* tests/data/test-read-dwarf/test1.abi: Likewise.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* tests/data/test-read-dwarf/test0.cc: Define a member function
out-of-line, outside of its namespace.
* tests/data/test-read-dwarf/test0.abi: Update the .abi file.
* tests/data/test-read-dwarf/test0: Update the resulting binary.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-reader.h (read_corpus_from_native_xml)
(read_corpus_from_native_xml_file): Declare new entry points.
* include/abg-writer.h (write_corpus_to_native_xml)
(write_corpus_to_native_xml_file): Likewise.
* src/abg-reader.cc (read_translation_unit_from_input): Renamed
read_input into this. Support new 'path' attribute for
'abi-instr' XML element.
(read_corpus_from_input): New static function.
(read_translation_unit_from_file)
(read_translation_unit_from_buffer)
(read_translation_unit_from_istream): Update wrt read_input ->
read_translation_unit_from_input.
(read_corpus_from_native_xml, read_corpus_from_native_xml)
(read_corpus_from_native_xml_file): Define new entry points.
* src/abg-writer.cc (write_translation_unit): Write 'path'
attribute into the 'abi-instr' xml element.
(write_corpus_to_native_xml, write_corpus_to_native_xml_file):
Define new entry points.
* tools/abg-tools-utils.h (file_type::{FILE_TYPE_XML_CORPUS,
FILE_TYPE_ZIP_CORPUS}): New enumerators.
* tools/abg-tools-utils.cc (guess_file_type): Support detection of
the new xml file format containing a document root 'abi-corpus'
root element.
* tools/bidiff.cc (main): Support diffing xml corpus-es and zip
corpus-es.
* tools/bidw.cc (main): Recognize elf files before reading them.
* tools/bilint.cc (main): Support reading xml/zip corpus-es too.
* tests/data/test-read-write/test[0-23].xml: Update 'path'
attribute.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-ir.h (ir_traversable_base): New type to be the base
of IR nodes that are to be traversed. Extends traversable_base.
Its ir_traversable_base::traversable() method takes an
ir_node_visitor&.
(ir_node_visitor::visit): Change these virtual overloads to take
pointers to IR nodes, rather than references. This will be useful
to e.g, store these IR nodes in containers on the side for some
algorithms to work. That is going to be useful later to,
e.g. build symbol tables on the side, using the visitor interface.
(class decl_base): Make this inherit ir_traversable_base.
* src/abg-ir.cc (*::traverse): Adjust comments and the call the
ir_node_visitor::visit call. Use the ir_traversable_base type
rather than traversable_base.
(ir_traversable_base::traverse): Define.
(ir_node_visitor::visit): Change these overloads to take pointers
rather than reference to ir nodes.
* tests/test-walker.cc (name_printing_visitor::visit): Adjust to
take pointers rather than references.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-ir.h (translation_unit::translation_unit): Take an
address_size parameter.
(translation_unit::{get_address_size, set_address_size}): New
accessors.
* src/abg-ir.cc (translation_unit::priv::address_size_): New
private member.
(translation_unit::translation_unit): Take an address_size
parameter.
(translation_unit::{get_address_size, set_address_size}): Define
these new methods.
* src/abg-reader.cc (read_input): Read the "address-size"
attribute from the abi-instr element.
* src/abg-writer.cc (write_translation_unit): Write the
"address-size" attribute.
* tests/data/test-read-write/test23.xml: New test input.
* tests/test-read-write.cc (in_out_specs): Add the new test above
to the list of files to read and write back.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/abg-reader.cc (build_function_parameter): Support reading
the "is-variadic" attribute.
* src/abg-writer.cc (write_function_decl): Support writing the
"is-variadic" attribute.
* tests/data/test-read-write/test22.xml: New test input.
* tests/test-read-write.cc (in_out_specs): Add the new test to the
list of files to read and write back.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-reader.h: New file with abigail::xml_reader APIs.
* include/abg-writer.h: New file with abigail::xml_writer APIs.
* include/Makefile.am: Add the new files above to the source
distribution.
* src/abg-reader.cc: Update top-file comments.
(namespace xml_reader): Rename namespace
reader into this.
(read_to_translation_unit, read_corpus_from_archive): New static
functions.
(read_translation_unit_from_file)
(read_translation_unit_from_file, read_corpus_from_file): New
entry points.
(struct array_deleter): New functor.
(translation_unit::read): Remove this.
* src/abg-writer.cc: Update top file comments.
(namespace xml_writer): Rename namespace
writer into this.
(struct archive_write_ctxt): New internal type.
(create_archive_write_context, write_translation_unit_to_archive)
(write_translation_unit, write_corpus_to_archive): New low level
static functions overloads.
(write_corpus_to_archive, write_translation_unit): Public higher
level overloads.
(translation_unit::write): Remove.
(dump): Update for new xml_writer namespace.
* include/abg-ir.h (translation_unit::{read, write}): Remove these
serialization methods.
* include/abg-corpus.h (corpus_sptr): New convenience typedef.
(corpus::{read, write}): Remove these methods.
* src/abg-corpus.cc (corpus::{read, write})
(corpus::impl::{serialized_tus, archive}): Remove these members.
(corpus::impl::{get_archive, close_archive, write_tu_to_archive,
read_to_translation_unit}): Remove these methods.
* tests/test-bidiff.cc (main): Update for usage of the new
xml_reader API.
* tests/test-read-write.cc (main): Likewise. Update for the usage
of the new xml_writer API, too.
* tests/test-walker.cc (main): Update for the usage of the new
xml_reader API.
* tests/test-write-read-archive.cc (main): Likewise. And for the
xml_writer API, too.
* tools/biar.cc (add_tus_to_archive, extract_tus_from_archive): Likewise.
* tools/bidiff.cc (main): Likewise, for xml_reader APIs.
* tools/bilint.cc (main): Likewise, for xml_writer APIs, too.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* tests/Makefile.am (data/test-read-write/test{17, 18, 19, 20,
21}.xml): Add these test input files to the source distribution.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-fwd.h (get_type_declaration): Declare function.
* include/abg-ir.h (class decl_base): Add class_decl as a friend.
This to be able to call decl_base::set_scope from class_decl.
(scope_decl::add_member_decl): Make this virtual protected, so
that it can be called (virtually) from e.g, class_decl.
(type_decl_sptr, typedef_decl_sptr): New convenience typedefs.
(class_decl::add_member_decl): New virtual overload for
scope_decl::add_member_decl.
(class_decl::{add_member_type, add_data_member,
add_member_function}): New overloads.
* src/abg-ir.cc (add_decl_to_scope): Benign style cleanup.
(get_type_declaration): Define new function.
(class_decl::add_member_decl): New method.
(class_decl::add_member_type): Avoid silently added a new member
type when that member type has already been (perhaps
inadvertently) added to a scope already. Rather, put a strict
assert in place there. Also add a new overload that constructs
the member type out of a classic type and adds it to the class.
(class_decl::{add_data_member, add_member_function}): Likewise.
(class_decl::{add_member_function_template,
add_member_class_template}): Avoid silently added a new member
template when that template has already been (perhaps
inadvertently) added to a scope already. Rather, put a strict
assert in place there.
* src/abg-reader.cc (push_decl_to_current_scope): Take a an extra
flag saying if the current decl should be added to the current
scope as well (in addition to being pushed onto the stack of
scopes maintained in the reader context).
(push_and_key_type_decl): Likewise, take that extra flag and pass
it to push_decl_to_current_scope.
(build_function_decl, build_var_decl, build_type_decl)
(build_qualified_type_decl, build_pointer_type_def)
(build_reference_type_def, build_enum_type_decl, build_typedef_decl)
(build_function_tdecl, build_class_tdecl): Likewise.
(build_class_decl): Likewise. When building member data, types,
and functions, make sure /not/ to add the data, type of function to
the current scope before adding it to the class_decl. This was
making the member not being added to the class because it already
had a scope.
(build_type_tparameter, build_type_composition)
(build_non_type_tparameter, build_template_tparameter)
(build_type): Adjust to add the template parm to the current scope
explicitly, like previously.
(handle_type_decl): Use build_type_decl function. Add the
type_decl to the current scope, like previously.
(handle_namespace_decl, handle_qualified_type_decl)
(handle_pointer_type_def, handle_reference_type_def)
(handle_enum_type_decl, handle_typedef_decl, handle_var_decl)
(handle_function_decl, handle_class_decl, handle_function_tdecl)
(handle_class_tdecl): Adjust to add the decl to the current scope,
like previously.
* tests/data/test-read-write/test21.xml: New test input with
member type(def).
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-diff-utils.h (point::set): New overload..
(point::{add, operator<, operator>, operator<=, operator>=}): New
methods.
(point::operator!=): Constify.
(point::operator==): Constify. Cleanup.
(point::operator=): Keep emptiness.
(class snake): New class definition
(d_path_vec::{over_bounds, offset}): New methods.
(d_path_vec::check_index_against_bound): Don't take a bound
parameter anymore. Use the new over_bound method above. Fix up
error reporting.
(d_path_vec::d_path_vec): Fix d_path_vec size allocation.
(d_path_vec::operator[]): Use the d_path_vec::at method to check
all accesses against the bounds. This is slower, but at least we
can expect to have something that is more robust. We can remove
the bound checking later when we are sure the code has been tested
enough. Also use the new offset() method.
(d_path_vec::at): Take long long.
(ends_of_furthest_d_paths_overlap): Constify input parameters.
(end_of_fr_d_path_in_k, end_of_frr_d_path_in_k_plus_delta): Take
an instance of the new snake in parameter, rather than a bare end
point that wasn't carrying enough information about the snake.
Record the snake which consists of up to four points: a begin
point, an intermediate point, a diagonal start point and an end
point. Return that snake upon successful completion.
(compute_middle_snake): Take an instance of snake, rather than the
two points that were supposed to represent a snake and with which
we were loosing information before. Revisit/simplify the logic of
this function; this literally goes forward or in reverse, gets the
resulting snake returned by the end_of_fr_d_path_in_k and
end_of_frr_d_path_in_k_plus_delta functions, detect if these snakes
overlap and just return the current snake. Much simpler. The
caller now gets a snake, which has much more information than the
previous snake approximation made of just two points. Bonus
point, this follows almost to the word, what the paper says.
(maybe_record_match_point, find_snake_start_point): Remove these
as there are not used by compute_middle_snake anymore.
(print_snake, ses_len): Update these to take/handle a snake.
(snake_end_points): New declaration.
(compute_diff): When we are getting an empty first sequence, this
means that we are inserting the second sequence *before* the
beginning of the first sequence; keep this information by setting
the insertion point index to -1, rather than zero. Update this to
get/handle snakes, rather than free points vaguely representing
snakes. Now that compute_middle_snake returns real snakes, handle
the information we are getting. Basically for edit scripts of
length equal to 1, as the snake carries all the necessary
information about the non-diagonal edge (as well as the diagonal
edges), we (can) now precisely update the current edit script (as
well as the longest common sub-sequence). For edit scripts of
length greater than 1, better at which points to divide the
problem and consequently, at which points to conquer it back --
better following The Paper to the letter.
(display_edit_script): Update this for the use of instances of
snake.
* src/abg-diff-utils.cc (ends_of_furthest_d_paths_overlap): Update
for constification of inputs.
(snake_end_points): Define new function.
(compute_middle_snake): Adapt for the taking an instance of snake.
* tests/test-diff2.cc (main): Update for using instances of snake.
* tests/test-core-diff.cc: Add new tests.
* tests/data/test-core-diff/report0.txt: Update for output
adaptation.
* tests/data/test-core-diff/report6.txt: Likewise.
* tests/data/test-core-diff/report7.txt: Likewise.
* tests/data/test-core-diff/report8.txt: New test data.
* tests/data/test-core-diff/report9.txt: Likewise.
* tests/data/test-core-diff/report10.txt: Likewise.
* tests/data/test-core-diff/report11.txt: Likewise.
* tests/data/test-core-diff/report12.txt: Likewise.
* tests/data/test-core-diff/report3.txt: Likewise.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-ir.h (type_shared_ptr_equal::operator()): Fix thinko
in checking for the boolean value of the pointers to types.
* src/abg-ir.cc (type_decl::operator==(const decl_base&)): Do not
forget to compare the decl_base part of the type too.
(type_decl::operator==(const type_base&)): To ease maintenance,
re-use the equality operator that takes a decl_base.
(scope_type_decl::operator==(const type_base&)): Likewise.
(qualified_type_def::operator==(const type_base&)): Likewise.
(compare_function_types): New sub-routine to compare function
types. It fixes an infinite recursion when comparing two methods
of the same class.
(function_type::operator==(const type_base&)): Use the new
compare_function_types function.
(class_decl::operator==(const decl_base&)): Fix a thinko in the
first test of the function. Use a dedicated scope for each class
section comparison; that way, there won't be any chance to misuse
the variables pertaining to a different section. Fix the member
function sections; we were mistakenly using the variables for the
*data* section there.
(class_decl::operator==(const type_base&)): Re-use the operator
that takes a decl_base.
(class_decl::operator==(const class_decl&)): Don't remove
const-ness during the static cast.
(class_decl::member_function::operator==(const member_function&)):
Do not remove the reference from the static cast.
(class_decl::member_class_template::operator==(const
member_base&)): Likewise.
(type_tparameter::operator==(const template_parameter&)):
Likewise.
(template_tparameter::operator==(const template_parameter&)):
Likewise.
(function_tdecl::operator==(const template_decl&)): Likewise.
(class_tdecl::operator==(const template_decl&)): Likewise.
(class_tdecl::operator==(const class_tdecl&)): Likewise.
* tests/data/test-read-write/test12.xml: Update this because the
test now correctly considers two type template parameters at the
same index as being equivalent.
* tests/data/test-read-write/test13.xml: Likewise.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-diff-utils.h (compute_middle_snake): After the
overlap determination happened, finding the middle snake can
require keep on building the current path until the "end". The
end meaning reaching the max of D. And that max is (M + N)/2 + 1.
In the extreme cases were middle snake was on the very last step
(M + N) + 1, we were not finding the middle snake. Fix this.
(compute_diff): When d == 1 and the first edge on the edit graph
is a non-diagonal edge and when a_base != a_begin, we were failing
to properly initialize x,y to find that non-diagonal edge. Also
we were failing to correctly compute the size of the sequence.
Fix these.
* tests/test-core-diff.cc: Add a new regression test for the two
cases above.
* tests/data/test-core-diff/report7.txt: New reference data for
the new regression test.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-diff-utils.h (point::{operator!=,operator==}): New
operators.
(end_of_fr_d_path_in_k, end_of_frr_d_path_in_k_plus_delta): Allow
the initial point (-1,-1) that is not a point addressing elements
of the input sequences, but that is the starting point of the
forward paths and the ending point of reverse paths in the "Linear
Refinement" of the algorithm.
(is_match_point, maybe_record_match_point)
(find_snake_start_point): New functions.
(find_last_snake_in_path): Remove this. It's not used anymore.
(compute_middle_snake): Allow checking for overlapping paths even
on points that are outside of the edit graph boundaries. Once the
overlap is detected, if a non-empty snake has been seen already,
report it as the middle snake. Otherwise, keep building the path
until the end and report the last snake encountered as the middle
snake. Add comments.
(compute_diff): For the d == 1 case, fix the logic of the finding
the non-diagonal edge. Fix typos. Add comments.
(display_edit_script): Fix report glitches.
* tests/data/test-core-diff/report3.txt: Update as per the report
glitch above.
* tests/data/test-core-diff/report4.txt: Likewise.
* tests/data/test-core-diff/report5.txt: Likewise.
* tests/data/test-core-diff/report6.txt: New reference report for
a new test.
* tests/test-core-diff.cc: Add a new test for negative delta.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* tests/test-diff2.cc: New command line testing facility.
* tests/Makefile.am: Add this to the build system.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* tests/test-utils.h (is_dir, ensure_dir_path_created)
(ensure_parent_dir_created): Move these directories manipulation
utilities from here to ...
* tools/abg-tools-utils.h (is_dir, ensure_dir_path_created)
(ensure_parent_dir_created): ... here in this new file.
(dir_name, base_name): Declare these new functions.
* tests/test-utils.cc (is_dir, ensure_dir_path_created)
(ensure_parent_dir_created): Likewise, move these to ...
* tools/abg-tools-utils.cc (is_dir, ensure_dir_path_created)
(ensure_parent_dir_created): ... here in this new file.
(dir_name, base_name): Define these.
* tools/Makefile.am: New file. Create a new libtoolsutils.la
static library with stuff from tools/abg-tools-utils.cc in it.
Also create a new 'biar' program with the stuff from the new
tools/biar.cc in it.
* tools/biar.cc: New file. Contains the code for the new "biar"
archive manipulation command line utility.
* tests/test-read-write.cc (main): Adjust for the change about
ensure_parent_dir_created above.
* tests/test-write-read-archive.cc (main): Likewise.
* Makefile.am (SUBDIRS): Add the new tools/ sub-directory to the
build system.
* configure.ac (AC_CONFIG_FILES): Generate tools/Makefile.
* tests/Makefile.am: Make libtestutils.la link with the new
libtoolsutils.la. Make sure to express the dependencies between
libtestutils.la and the binaries that depend on it. Otherwise
parallel builds can go awry.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* include/abg-corpus.h (corpus::{get_path, set_path): Renamed
corpus::get_file_path and corpus::set_file_path into these as
get_path/set_path is what is used elsewhere as well.
* src/abg-corpus.cc (corpus::{get_path, set_path}): Likewise.
* tests/test-write-read-archive.cc (main): Adjust for the change
above.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* configure.ac: Support detection of libzip dependency. Define
new DEPS_CFLAGS and DEPS_LIBS variables for use in
Makefile.am to refer to the dependency headers and
libraries.
* doc/website/mainpage.txt: Update this to talk about the new
libzip dependency.
* include/Makefile.am: Add abg-libzip-utils.h to the build system.
* include/abg-corpus.h (corps): Hide abigail::corpus's private behind a
pimpl idiom.
(corpus::{drop_translation_units, get_file_path, set_file_path,
write, read}): New methods.
* include/abg-libxml-utils.h (new_reader_from_buffer): Declare new
function.
* include/abg-libzip-utils.h: New file.
* src/Makefile.am: Add abg-corpus.cc and abg-libzip-utils.cc to
the build system. Refer to the library and headers dependencies
via the new DEPS_LIBS and DEPS_CFLAGS variables.
* src/abg-corpus.cc: New file.
* src/abg-ir.cc (translation::set_path): New method.
* src/abg-libxml-utils.cc (new_reader_from_buffer): Define new
function.
* src/abg-libzip-utils.cc: New file.
* src/abg-reader.cc (translation_unit::read): New overload.
* src/abg-writer.cc: Inject the names from the std namespace into
the abigail namespace, rather than into abigail::writer.
(abigail::translation_unit::write): New overload. This can now
use ofstream and the other stuff from std that are injected in the
abigail:: namespace.
* tests/Makefile.am: Add tests/test-write-read-archive.cc to the
build system; use that to build runtestwritereadarchive. Also add
the input test data from
tests/data/test-write-read-archive/test[0-4].xml.
* /tests/data/test-write-read-archive/test[0-4].xml: New test
input data files.
* tests/test-write-read-archive.cc: New test for this archive
write/read support.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* src/Makefile.am: Pass absolute file paths to the compiler during
the build. This helps in e.g in emacs' compilation mode, when the
output shows an error reported by GCC's diagnostics, setting point
to the error line and hitting 'enter' transports the user to the
file location where the error happened; as the file path is nows
absolute, emacs can always find it. Otherwise, finding it depends
on $PWD and whatnot.
* tests/Makefile.am: Likewise.
Signed-off-by: Dodji Seketeli <dodji@redhat.com>
* Makefile.in: Remove this autotools-generated file. I know that
people in the GCC-realm like storing these autotools-generated
files into the source control system, but I believe in this day
and age, this is annoying (to say the least) for the project
developers for no good reason. Requiring (the ubiquitous)
autoconf for building from source is no big deal today, and typing
"autoreconf" is not hard to do, really. And it saves the
developers for having to update a bunch of boilerplate
automatically generated files into the source control system; this
is really unnecessary noise and it is a commonly accepted good
practice to avoid doing it these days. To help people who do not
know how to handle this, there is a COMPILING file in the source
tree that explains how to build the project from sources.
* aclocal.m4: Likewise.
* configure: Likewise.
* include/Makefile.in: Likewise.
* src/Makefile.in: Likewise.
* doc/Makefile.in: Likewise.
* tests/Makefile.in: Likewise.
* include/libabigail/abg-ir.h (struct ir_node_visitor, struct
traversable): New interfaces.
(translation_unit, scope_decl, type_decl, qualified_type_def)
(pointer_type_def, reference_type_def, enum_type_decl)
(typedef_decl, var_decl, function_decl, data_member)
(member_function, member_function_template)
(member_class_template): Implement the traversable interface,
overload the traversable::traverse pure virtual function.
* src/abg-ir.cc ({translation_unit, scope_decl, type_decl,
namespace_decl, qualified_type_def, pointer_type_def,
reference_type_def, enum_type_decl, typedef_decl, var_decl,
function_decl, class_decl::member_function, class_decl,
class_decl::data_member, class_decl::member_function_template,
class_decl::member_class_template, function_template_decl,
class_template_decl, }::traverse): Implement traversal.
(ir_node_visitor::visit): New method, overloaded for the types
above, which implement the traversable interface.
* tests/test-walker.cc: New test case program to showcase how to
use the new traversal API.
* tests/makefile.am: Add test-walker.cc to the build system.
* include/libabigail: New directory.
* include/Makefile.am: New file.
* include/libabigail/Makefile.am: New file.
* src/abg-*.h: Move these in include/libabigail/*.h
* src/Makefile.am: Set -I option to look for headers in include/libabigail
* doc/api/libabigail.doxy: Look for headers in include/libabigail
* tests/Makefile.am: Set -I option to look for headers in include/libabigail
* abigail.m4: Set includedir to $prefix/include/libabigail for
library used as a dep, or to srcdir/libabigail/include/libabigail
in GCC source tree.
* configure.ac: Add Makefile.am in include and include/libabigail
* abigail.m4: Look for headers in include/libabigail when the
libabigail is in in the source tree and in $incdir/libabigail when
it is installed as a dependency.
* src/abg-ir.h (class_decl::base_spec::base_spec): Take an offset
and a is_virtual flag.
(class_decl::base_spec::{get_is_virtual, get_offset_in_bits}): New
methods.
(class_decl::has_no_base_nor_member): New method declaration..
* src/abg-ir.cc (class_decl::base_spec::base_spec): Take an offset
and a is_virtual flag
(class_decl::has_no_base_nor_member): Define it.
* src/abg-reader.cc (read_offset_in_bits): Renamed
read_var_offset_in_bits into this.
(read_is_virtual): New static function.
(build_class_decl): Read the 'layout-offset-in-bits' and the
'is-virtual' of the base class specifier. Adjust for the
read_var_offset_in_bits -> read_offset_in_bits rename.
* src/abg-writer.cc (write_layout_offset): New overload for the
base class specifiers.
(write_class_decl): If the class has no member or base class, make
it a proper empty xml element. Write the offset and the
is-virtual attribute.
* tests/data/test-read-write/test20.xml: New test input data.
* tests/test-read-write.cc: De-serialize the test above, serialize
it back and compare that both versions are the same.
* src/abg-ir.h (class_decl::base_spec::base_spec): Move this
out-of-line. Add a new constructor for clients compiled without
RTTI.
* src/abg-ir.cc (class_decl::base_spec::base_spec): This is now
here out-of-line. Define the new constructor for clients compiled
without RTTI.
* tests/data/test-read-write/test19.xml: New test case input for
base classes.
* tests/test-read-write.cc: De-serialize and serialize the new
test case input above.
* abg-ir.h (class_decl::set_earlier_declaration): Move this
out-of-line and add an overload.
* src/abg-ir.cc (class_decl::set_earlier_declaration): Moved this
out-of-line here and add an overload.
* src/abg-reader.cc (build_class_decl): Really read the
'decl-of-decl-id' property. Do not make the definition use the id
of the declaration. Both have their id.
* src/abg-writer.cc (write_class_decl): Put a space before the
property "def-of-decl-id".
* tests/data/test-read-write/test18.xml: New test input.
* tests/test-read-write.cc: De-serialize this new input, serialize
it back and make sure both versions are identical.
* src/abg-ir.h (function_decl::parameter::parameter): New
constructor with variadic parameter marker.
(function_decl::m_type): Make this protected to let method_decl
inheriting class to access it.
(function_decl::get_type): Move this out-of-line.
(class method_type, method_type_hash): New types.
(enum class_decl::access_specifier): Add no_access new enumerator.
(class_decl::data_member::data_member): Move this out-of-line.
(class_decl::data_member::~data_member): Declare virtual
destructor.
(class method_decl): New class.
(class member_function): Make this inherit method_decl, instead of
function_decl.
(class_decl::class_decl): New constructors.
(class_decl::{hashing_started, is_declaration_only,
set_earlier_declaration, get_earlier_declaration}): New methods.
* src/abg-ir.cc (add_decl_to_scope): If a decl is already in a
scope, don't add it to this scope.
(get_global_scope): Make this work when passed an instance of
global_scope.
(dynamic_type_hash::operator()): Add support for method_type.
(method_type::{method_type, set_class_type, ~method_type, })
(method_type_hash::operator()): New defintions.
(function_decl::get_type, class_decl::class_decl): Move these
out-of-line here.
(class_decl::method_decl::{method_decl, ~method_decl, get_type}):
New definitions.
(class_decl::member_function::member_function): Move this
out-of-line here. Support method_decl.
(class_decl::data_member::data_member): Likewise.
(class_decl_hash::operator()): Guard this against endless loop.
* src/abg-reader.cc (write_class_is_declaration_only): New static
function.
(write_var_decl): Take a flag to write the mangled name or not.
(write_function_decl): Take a flag to skip the first parameter.
(write_cdtor_const_static): Use 'yes' instead of 'true' as value
of the properties.
(write_decl, write_function_template_decl): Adjust wrt the new
signatures of write_var_decl and write_function_decl.
(write_enum_type_decl): Simplify call to write_location.
(write_class_decl): Support serializing declaration-only classes.
* src/abg-writer.cc:
* tests/data/test-read-write/test17.xml: New test input.
* tests/test-read-write.cc: De-serialize the above, and serialize it back.
* tests/data/test-read-write/test10.xml: Update this test.
* src/abg-ir.h (class function_type): Forward decl prior to class
function_decl.
(function_decl::parameter): Pass string by const reference. Add a
variadic marker member and initialize it.
(function_decl::parameter::get_type): Add a non-const overload.
(function_decl::parameter::get_variadic_marker): New getter.
(function_decl::function_decl) Take a const reference to a vector
for parameters, type size/alignment. Add two overloads that takes a
pointer to function_type.
(function_decl::get_parameters): Move this out-of-line.
(function_decl::append_parameter(s)): Renamed
function_decl::add_parameter(s) into these. Move it out-of-line.
Add an overload.
(function_decl::{get_type, set_type}): New declaration.
(function_decl::get_return_type): Move this out-of-line.
(function_decl::is_variadic): New in-line function.
(function_decl::m_type): New data member.
(function_decl::{m_parms, m_return_type}): Remove. This are now
carried by function_decl::m_type.
(class function_type, struct function_type_hash): New
declarations.
(member_function::member_function): Take a vector of pointers to
parameters. Take size/align of the type of the member function.
Adjust initialization.
* src/abg-ir.cc (dynamic_type_hash): Hash instance of
function_type accessed through a pointer.
(function_type::{operator==, ~function_type})
(function_type_hash::operaror(), function_decl::{get_return_type,
}, function_decl::parameter:#️⃣:operator()): New definitions.
(function_decl::function_decl): The out-of-line definitions of the
declarations above.
(function_decl::append_parameter): Moved this out-of-line from
inline function_decl::add_parameter. Make this rely on the
underlying m_type.
(function_decl::operator==): Adjust for use of vector for the
parameters. Also, there is no need anymore to compare the
parameters or the return types as they are compared by the
comparison of the function types.
* src/abg-reader.cc (build_function_decl): Read the new size/alignment
attributes on the function-decl element. Build a function_type
and use it to build the function_decl. Parameters and return type
are now hung off of the function_type.
(handle_function_decl): use build_function_decl.
* src/abg-writer.cc (write_function_decl): Write the new
size/alignment properties of the function-decl element. Adjust
for the use of vectors for function parameters now.
* tests/data/test-read-write/test10.xml: Adjust for the presence
of size/alignment properties in the function-decl element now.
* tests/data/test-read-write/test11.xml: Likewise.
* tests/data/test-read-write/test12.xml: Likewise.
* tests/data/test-read-write/test13.xml: Likewise.
* tests/data/test-read-write/test14.xml: Likewise.
* tests/data/test-read-write/test9.xml: Likewise.
