Improve randomized edits in corpus specs
This commit is contained in:
Max Brunsfeld
parent
dc28a10bf6
commit
2c2c567a29
1 changed files with 128 additions and 78 deletions
|
|
@ -1,5 +1,7 @@
|
|||
#include "runtime/runtime_spec_helper.h"
|
||||
#include <functional>
|
||||
#include <set>
|
||||
#include <utility>
|
||||
#include "runtime/length.h"
|
||||
#include "runtime/helpers/read_test_entries.h"
|
||||
#include "runtime/helpers/spy_input.h"
|
||||
|
|
@ -13,44 +15,6 @@ extern "C" const TSLanguage *ts_language_golang();
|
|||
extern "C" const TSLanguage *ts_language_c();
|
||||
extern "C" const TSLanguage *ts_language_cpp();
|
||||
|
||||
void expect_the_correct_tree(TSNode node, TSDocument *doc, string tree_string) {
|
||||
const char *node_string = ts_node_string(node, doc);
|
||||
AssertThat(node_string, Equals(tree_string));
|
||||
free((void *)node_string);
|
||||
}
|
||||
|
||||
void expect_a_consistent_tree(TSNode node, TSDocument *doc) {
|
||||
size_t child_count = ts_node_child_count(node);
|
||||
size_t start = ts_node_start_char(node);
|
||||
size_t end = ts_node_end_char(node);
|
||||
TSPoint start_point = ts_node_start_point(node);
|
||||
TSPoint end_point = ts_node_end_point(node);
|
||||
|
||||
bool has_changes = ts_node_has_changes(node);
|
||||
bool some_child_has_changes = false;
|
||||
|
||||
for (size_t i = 0; i < child_count; i++) {
|
||||
TSNode child = ts_node_child(node, i);
|
||||
size_t child_start = ts_node_start_char(child);
|
||||
size_t child_end = ts_node_end_char(child);
|
||||
TSPoint child_start_point = ts_node_start_point(child);
|
||||
TSPoint child_end_point = ts_node_end_point(child);
|
||||
|
||||
AssertThat(child_start, IsGreaterThan(start) || Equals(start));
|
||||
AssertThat(child_end, IsLessThan(end) || Equals(end));
|
||||
AssertThat(child_start_point, IsGreaterThan(start_point) || Equals(start_point));
|
||||
AssertThat(child_end_point, IsLessThan(end_point) || Equals(end_point));
|
||||
|
||||
if (ts_node_has_changes(child))
|
||||
some_child_has_changes = true;
|
||||
}
|
||||
|
||||
if (child_count > 0)
|
||||
AssertThat(has_changes, Equals(some_child_has_changes));
|
||||
}
|
||||
|
||||
START_TEST
|
||||
|
||||
map<string, const TSLanguage *> languages({
|
||||
{"json", ts_language_json()},
|
||||
{"arithmetic", ts_language_arithmetic()},
|
||||
|
|
@ -60,22 +24,100 @@ map<string, const TSLanguage *> languages({
|
|||
{"cpp", ts_language_cpp()},
|
||||
});
|
||||
|
||||
void expect_the_correct_tree(TSNode node, TSDocument *doc, string tree_string) {
|
||||
const char *node_string = ts_node_string(node, doc);
|
||||
AssertThat(node_string, Equals(tree_string));
|
||||
free((void *)node_string);
|
||||
}
|
||||
|
||||
void expect_a_consistent_tree(TSNode node, TSDocument *doc) {
|
||||
size_t child_count = ts_node_child_count(node);
|
||||
size_t start_char = ts_node_start_char(node);
|
||||
size_t end_char = ts_node_end_char(node);
|
||||
TSPoint start_point = ts_node_start_point(node);
|
||||
TSPoint end_point = ts_node_end_point(node);
|
||||
bool has_changes = ts_node_has_changes(node);
|
||||
bool some_child_has_changes = false;
|
||||
|
||||
AssertThat(start_char, !IsGreaterThan(end_char));
|
||||
AssertThat(start_point, !IsGreaterThan(end_point));
|
||||
|
||||
size_t last_child_end_char = 0;
|
||||
TSPoint last_child_end_point = {0, 0};
|
||||
|
||||
for (size_t i = 0; i < child_count; i++) {
|
||||
TSNode child = ts_node_child(node, i);
|
||||
size_t child_start_char = ts_node_start_char(child);
|
||||
size_t child_end_char = ts_node_end_char(child);
|
||||
TSPoint child_start_point = ts_node_start_point(child);
|
||||
TSPoint child_end_point = ts_node_end_point(child);
|
||||
|
||||
if (i > 0) {
|
||||
AssertThat(child_start_char, !IsLessThan(last_child_end_char));
|
||||
AssertThat(child_start_point, !IsLessThan(last_child_end_point));
|
||||
last_child_end_char = child_end_char;
|
||||
last_child_end_point = child_end_point;
|
||||
}
|
||||
|
||||
AssertThat(child_start_char, !IsLessThan(start_char));
|
||||
AssertThat(child_end_char, !IsGreaterThan(end_char));
|
||||
AssertThat(child_start_point, !IsLessThan(start_point));
|
||||
AssertThat(child_end_point, !IsGreaterThan(end_point));
|
||||
|
||||
expect_a_consistent_tree(child, doc);
|
||||
|
||||
if (ts_node_has_changes(child))
|
||||
some_child_has_changes = true;
|
||||
}
|
||||
|
||||
if (child_count > 0)
|
||||
AssertThat(has_changes, Equals(some_child_has_changes));
|
||||
}
|
||||
|
||||
string random_string(char min, char max) {
|
||||
string result;
|
||||
size_t length = rand() % 12;
|
||||
for (size_t i = 0; i < length; i++) {
|
||||
char inserted_char = min + (rand() % (max - min));
|
||||
result += inserted_char;
|
||||
}
|
||||
return result;
|
||||
}
|
||||
|
||||
string random_char(string characters) {
|
||||
size_t index = rand() % characters.size();
|
||||
return string() + characters[index];
|
||||
}
|
||||
|
||||
string random_words(size_t count) {
|
||||
string result;
|
||||
for (size_t i = 0; i < count; i++) {
|
||||
if (!result.empty() && rand() % 2 > 0)
|
||||
result += " ";
|
||||
if (rand() % 10 < 5)
|
||||
result += random_char("!(){}[]<>+-=");
|
||||
else
|
||||
result += random_string('a', 'z');
|
||||
}
|
||||
return result;
|
||||
}
|
||||
|
||||
START_TEST
|
||||
|
||||
describe("Languages", [&]() {
|
||||
TSDocument *doc;
|
||||
|
||||
before_each([&]() {
|
||||
doc = ts_document_make();
|
||||
});
|
||||
|
||||
after_each([&]() {
|
||||
ts_document_free(doc);
|
||||
});
|
||||
srand(0);
|
||||
|
||||
for (const auto &pair : languages) {
|
||||
describe(("The " + pair.first + " parser").c_str(), [&]() {
|
||||
TSDocument *doc;
|
||||
|
||||
before_each([&]() {
|
||||
doc = ts_document_make();
|
||||
ts_document_set_language(doc, pair.second);
|
||||
// ts_document_set_debugger(doc, log_debugger_make(false));
|
||||
});
|
||||
|
||||
after_each([&]() {
|
||||
ts_document_free(doc);
|
||||
});
|
||||
|
||||
for (auto &entry : test_entries_for_language(pair.first)) {
|
||||
|
|
@ -97,49 +139,57 @@ describe("Languages", [&]() {
|
|||
ts_document_parse(doc);
|
||||
});
|
||||
|
||||
srand(2);
|
||||
std::set<std::pair<size_t, size_t>> deletions;
|
||||
std::set<std::pair<size_t, string>> insertions;
|
||||
|
||||
for (int i = 0; i < 5; i++) {
|
||||
for (size_t i = 0; i < 50; i++) {
|
||||
size_t edit_position = rand() % entry.input.size();
|
||||
size_t deletion_amount = rand() % (entry.input.size() - edit_position);
|
||||
string pos_string = to_string(edit_position);
|
||||
size_t deletion_size = rand() % (entry.input.size() - edit_position);
|
||||
string inserted_text = random_words(rand() % 4 + 1);
|
||||
|
||||
it_handles_edit_sequence("repairing an inserted error at " + pos_string, [&]() {
|
||||
ts_document_edit(doc, input->replace(edit_position, 0, "%^&*"));
|
||||
ts_document_parse(doc);
|
||||
if (insertions.insert({edit_position, inserted_text}).second) {
|
||||
string description = "'" + inserted_text + "' at " + to_string(edit_position);
|
||||
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
it_handles_edit_sequence("repairing an insertion of " + description, [&]() {
|
||||
ts_document_edit(doc, input->replace(edit_position, 0, inserted_text));
|
||||
ts_document_parse(doc);
|
||||
|
||||
it_handles_edit_sequence("creating and repairing an inserted error at " + pos_string, [&]() {
|
||||
ts_document_parse(doc);
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
|
||||
ts_document_edit(doc, input->replace(edit_position, 0, "%^&*"));
|
||||
it_handles_edit_sequence("performing and repairing an insertion of " + description, [&]() {
|
||||
ts_document_parse(doc);
|
||||
|
||||
ts_document_parse(doc);
|
||||
ts_document_edit(doc, input->replace(edit_position, 0, inserted_text));
|
||||
ts_document_parse(doc);
|
||||
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
}
|
||||
|
||||
it_handles_edit_sequence("repairing an errant deletion at " + pos_string, [&]() {
|
||||
ts_document_parse(doc);
|
||||
if (deletions.insert({edit_position, deletion_size}).second) {
|
||||
string desription = to_string(edit_position) + "-" + to_string(edit_position + deletion_size);
|
||||
|
||||
ts_document_edit(doc, input->replace(edit_position, deletion_amount, ""));
|
||||
ts_document_parse(doc);
|
||||
it_handles_edit_sequence("repairing a deletion of " + desription, [&]() {
|
||||
ts_document_edit(doc, input->replace(edit_position, deletion_size, ""));
|
||||
ts_document_parse(doc);
|
||||
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
|
||||
it_handles_edit_sequence("creating and repairing an errant deletion at " + pos_string, [&]() {
|
||||
ts_document_edit(doc, input->replace(edit_position, deletion_amount, ""));
|
||||
ts_document_parse(doc);
|
||||
it_handles_edit_sequence("performing and repairing a deletion of " + desription, [&]() {
|
||||
ts_document_parse(doc);
|
||||
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
ts_document_edit(doc, input->replace(edit_position, deletion_size, ""));
|
||||
ts_document_parse(doc);
|
||||
|
||||
ts_document_edit(doc, input->undo());
|
||||
ts_document_parse(doc);
|
||||
});
|
||||
}
|
||||
}
|
||||
}
|
||||
});
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue