diff options
| -rw-r--r-- | compiler/aup.h | 2 | ||||
| -rw-r--r-- | compiler/compile.c | 61 | ||||
| -rw-r--r-- | compiler/main.c | 27 | ||||
| -rw-r--r-- | tests/check.py | 16 |
4 files changed, 69 insertions, 37 deletions
diff --git a/compiler/aup.h b/compiler/aup.h index 98b59b0..c35c44a 100644 --- a/compiler/aup.h +++ b/compiler/aup.h @@ -4,7 +4,7 @@ #include <stddef.h> /* aup: source text to a Pure module (Pure docs/design/008-module-format.md). */ -enum { AUP_SOURCE_LIMIT = 32768, AUP_MESSAGE = 256, AUP_PATH = 256 }; +enum { AUP_MESSAGE = 256, AUP_PATH = 256 }; typedef struct { char path[AUP_PATH]; /* the library the error is in, or "" for the main source */ diff --git a/compiler/compile.c b/compiler/compile.c index cdb059a..aad75de 100644 --- a/compiler/compile.c +++ b/compiler/compile.c @@ -8,11 +8,13 @@ #include <stdlib.h> #include <string.h> -/* Limits and opcodes of Pure module format v6. */ -enum { STATES = 32, INPUTS = 32, VIEWS = 64, HANDLERS = 64, CODE = 65535, STACK = 64, NAME = 32, JOINS = 8, +/* Limits and opcodes of Pure module format v7. Pure counts its states, inputs, views, handlers, records and record fields + * in u16s alone (027's 4f): the compiler's tables hold these many (their rooms unused but by what a program declares). A + * message's fields (FIELDS) and a record's (FIELDS_PER_RECORD) are Pure's widths; functions Pure's 64. */ +enum { STATES = 4096, INPUTS = 1024, VIEWS = 4096, HANDLERS = 4096, CODE = 65535, STACK = 64, NAME = 32, JOINS = 8, DEPTH = 64, FIELDS = 64, CONSTANTS = 4096, CONSTANT_BYTES = 1 << 20, - RECORDS = 16, RECORD_FIELDS = 128, FIELDS_PER_RECORD = 16, FNS = 64, PARAMS = 8, MODULES = 9, - LIBRARY_RECORDS = 64 }; + RECORDS = 1024, RECORD_FIELDS = 16384, FIELDS_PER_RECORD = 16, FNS = 64, PARAMS = 8, MODULES = 9, + LIBRARY_RECORDS = 1024, OUTPUTS = 255 }; enum { CONST, LOAD, ADD, SUB, MUL, DIV, NEG, LT, LE, GT, GE, EQ, NE, AND, OR, NOT, MIN, MAX, CLAMP, STORE, JZ, JMP, FIELD, LEN, REPLY, REQUIRE, TEXT_CONST, TEXT_LOAD, NUM_TEXT, CONCAT, CHARS, LINES, TEXT_BYTES, @@ -1740,7 +1742,7 @@ static void record_declaration(Compiler *c) { Token name = expect_name(c, "record name"); declare(c, &name); bool library = c->module != 0; /* a library's: into the module once the app holds it in a list */ - if (!library && c->record_count == RECORDS) fail_at(c, name.line, name.column, "at most 16 records are allowed"); + if (!library && c->record_count == RECORDS) fail_at(c, name.line, name.column, "at most %d records are allowed", RECORDS); if (library && c->library_record_count == LIBRARY_RECORDS) fail_at(c, name.line, name.column, "the libraries declare at most %d records", LIBRARY_RECORDS); Record *r = library ? &c->library_records[c->library_record_count] : &c->records[c->record_count]; @@ -1791,7 +1793,7 @@ static void record_declaration(Compiler *c) { if (!r->typed) { /* plain: its fields as written */ for (unsigned i = 0; i < r->declared_count; ++i) { if (!library && c->record_fields == RECORD_FIELDS) - fail_at(c, r->declared[i].line, r->declared[i].column, "at most 128 record fields are allowed"); + fail_at(c, r->declared[i].line, r->declared[i].column, "at most %d record fields are allowed", RECORD_FIELDS); memcpy(r->field_names[i], r->declared[i].name, NAME); r->field_kinds[i] = r->declared[i].kind; r->field_number[i] = r->declared[i].number; @@ -1811,7 +1813,7 @@ static void variant_declaration(Compiler *c) { Token name = expect_name(c, "variant name"); declare(c, &name); if (c->module) fail_at(c, name.line, name.column, "a library declares no variant, for now"); - if (c->record_count == RECORDS) fail_at(c, name.line, name.column, "at most 16 records are allowed, variants among them"); + if (c->record_count == RECORDS) fail_at(c, name.line, name.column, "at most %d records are allowed, variants among them", RECORDS); Record *r = &c->records[c->record_count]; memset(r, 0, sizeof(*r)); memcpy(r->name, name.text, NAME); @@ -1865,7 +1867,7 @@ static Token held_record(Compiler *c, char name[NAME], int *library) { static void state_declaration(Compiler *c, bool transient) { Token name = expect_name(c, "state name"); declare(c, &name); - if (c->states == STATES) fail_at(c, name.line, name.column, "at most 32 states are allowed"); + if (c->states == STATES) fail_at(c, name.line, name.column, "at most %d states are allowed", STATES); unsigned index = c->states; memcpy(c->state_names[index], name.text, NAME); c->state_transient[index] = transient; @@ -1952,7 +1954,7 @@ static void table_declaration(Compiler *c, unsigned table) { for (unsigned i = table + 1; i < c->views; ++i) if (!strcmp(c->view_names[i], column.text)) fail_at(c, column.line, column.column, "duplicate column '%s'", column.text); - if (c->views == VIEWS) fail_at(c, column.line, column.column, "at most 64 views are allowed"); + if (c->views == VIEWS) fail_at(c, column.line, column.column, "at most %d views are allowed", VIEWS); expect(c, "=", "'='"); unsigned index = c->views++; memcpy(c->view_names[index], column.text, NAME); @@ -2040,7 +2042,6 @@ static void use_declaration(Compiler *c) { char path[AUP_PATH] = ""; if (!c->load || !c->load(c->load_context, name.text, &source, &length, path)) fail_at(c, name.line, name.column, "cannot find library '%s' (%s.aup)", name.text, name.text); - if (length > AUP_SOURCE_LIMIT) fail_at(c, name.line, name.column, "library '%s' exceeds 32768 bytes", name.text); library = (int)c->module_count++; memcpy(c->modules[library].name, name.text, NAME); snprintf(c->modules[library].path, AUP_PATH, "%s", *path ? path : name.text); @@ -2088,7 +2089,7 @@ static void library_declaration(Compiler *c) { * input of their own named view.args, which a request gives (?name=value). */ static void view_arguments(Compiler *c, unsigned view, const Token *name) { take(c); /* ( */ - if (c->input_count == INPUTS) fail_at(c, name->line, name->column, "at most 32 inputs are allowed (a view's arguments take one)"); + if (c->input_count == INPUTS) fail_at(c, name->line, name->column, "at most %d inputs are allowed (a view's arguments take one)", INPUTS); Input *in = &c->inputs[c->input_count]; memset(in, 0, sizeof(*in)); if (strlen(name->text) + 5 >= NAME) @@ -2150,7 +2151,7 @@ static void declaration(Compiler *c) { } else if (!strcmp(kind.text, "input")) { Token name = expect_name(c, "input name"); declare(c, &name); - if (c->input_count == INPUTS) fail_at(c, name.line, name.column, "at most 32 inputs are allowed"); + if (c->input_count == INPUTS) fail_at(c, name.line, name.column, "at most %d inputs are allowed", INPUTS); expect(c, "=", "'='"); Input *in = &c->inputs[c->input_count]; memset(in, 0, sizeof(*in)); @@ -2206,7 +2207,10 @@ static void declaration(Compiler *c) { /* output ordered = message { ... }: what the app sends; its receiver is the host's wiring. */ Token name = expect_name(c, "output name"); declare(c, &name); - if (c->input_count == INPUTS) fail_at(c, name.line, name.column, "at most 32 inputs and outputs are allowed"); + if (c->input_count == INPUTS) fail_at(c, name.line, name.column, "at most %d inputs and outputs are allowed", INPUTS); + unsigned outputs = 0; /* Pure's store names them in a byte's count */ + for (unsigned i = 0; i < c->input_count; ++i) outputs += c->inputs[i].kind == KIND_OUTPUT; + if (outputs == OUTPUTS) fail_at(c, name.line, name.column, "at most %d outputs are allowed", OUTPUTS); expect(c, "=", "'='"); Input *in = &c->inputs[c->input_count]; memset(in, 0, sizeof(*in)); @@ -2222,7 +2226,7 @@ static void declaration(Compiler *c) { } else if (!strcmp(kind.text, "view")) { Token name = expect_name(c, "view name"); declare(c, &name); - if (c->views == VIEWS) fail_at(c, name.line, name.column, "at most 64 views are allowed"); + if (c->views == VIEWS) fail_at(c, name.line, name.column, "at most %d views are allowed", VIEWS); unsigned index = c->views++; if (is(c, "(")) view_arguments(c, index, &name); memcpy(c->view_names[index], name.text, NAME); @@ -2239,7 +2243,7 @@ static void declaration(Compiler *c) { else table_declaration(c, index); } } else if (!strcmp(kind.text, "on")) { - if (c->handler_count == HANDLERS) fail_at(c, kind.line, kind.column, "at most 64 handlers are allowed"); + if (c->handler_count == HANDLERS) fail_at(c, kind.line, kind.column, "at most %d handlers are allowed", HANDLERS); Token event = expect_name(c, "frame or an input"); unsigned index = c->handler_count++; c->handlers[index].line = event.line; @@ -2713,7 +2717,7 @@ static int hidden_list(Compiler *c, int record, bool bound, int line, int column for (unsigned i = 0; !bound && i < c->states; ++i) if (c->state_hidden[i] && !c->state_bound[i] && c->state_record[i] == record) return (int)i; if (c->states == STATES) - fail_at(c, line, column, "at most 32 states are allowed, the lists the code reads in records among them"); + fail_at(c, line, column, "at most %d states are allowed, the lists the code reads in records among them", STATES); unsigned i = c->states++; snprintf(c->state_names[i], NAME, "%s.r%u", bound ? "bound" : "empty", i); c->state_types[i] = STATE_LIST; @@ -3759,6 +3763,8 @@ static Kind value(Compiler *c, const Expr *e, bool handler, int height) { expect_kind(c, e->args[2], wanted(c, e->args[2], handler, height, key), key); expect_kind(c, e->args[3], wanted(c, e->args[3], handler, height + 1, key), key); } + if (list > 0xFF) /* a fold names its list in a byte of its argument (Pure's FOLD) */ + fail_at(c, e->line, e->column, "a folded list is among the first 256 states; '%s' is state %d", e->member, list); unsigned at = c->code_count, joins = c->joins; emit(c, e->line, e->column, FOLD, 0, 0); c->folding = true; @@ -4812,6 +4818,8 @@ static void statements(Compiler *c, const Stmt *s) { expect_kind(c, s->low, wanted(c, s->low, c->handling, 0, key), key); expect_kind(c, s->high, wanted(c, s->high, c->handling, 1, key), key); } + if (list > 0xFF) /* a walk names its list in a byte of its argument (Pure's EACH) */ + fail_at(c, s->line, s->column, "a walked list is among the first 256 states; '%s' is state %d", s->member, list); unsigned at = c->code_count, jump = 0, loop = c->loop_count++; emit(c, s->line, s->column, EACH, 0, 0); memcpy(c->loops[loop].name, s->name, NAME); @@ -5119,7 +5127,7 @@ static void event_lists(Compiler *c) { } continue; } - if (c->states == STATES) fail_at(c, s->line, s->column, "at most 32 states are allowed, a let's list among them"); + if (c->states == STATES) fail_at(c, s->line, s->column, "at most %d states are allowed, a let's list among them", STATES); unsigned i = c->states++; memcpy(c->state_names[i], path, strlen(path) + 1); c->state_types[i] = STATE_LIST; @@ -5144,7 +5152,7 @@ static void rows_lists(Compiler *c) { if (length < 0 || length >= NAME) fail_at(c, in->field_line[f], in->field_column[f], "'%s' is longer than %d characters", name, NAME - 1); if (c->states == STATES) - fail_at(c, in->field_line[f], in->field_column[f], "at most 32 states are allowed (a message's rows take one)"); + fail_at(c, in->field_line[f], in->field_column[f], "at most %d states are allowed (a message's rows take one)", STATES); unsigned index = c->states++; memcpy(c->state_names[index], name, (size_t)length + 1); memcpy(c->state_target[index], in->field_record[f], NAME); @@ -5181,9 +5189,9 @@ static void pull_records(Compiler *c) { fail_at(c, r->used_line, r->used_column, "%s.%s is '%s' in the module, and '%s' is %s here", library, r->name, r->name, r->name, taken); if (c->record_count == RECORDS) - fail_at(c, r->used_line, r->used_column, "at most 16 records are allowed (%s.%s is one more)", library, r->name); + fail_at(c, r->used_line, r->used_column, "at most %d records are allowed (%s.%s is one more)", RECORDS, library, r->name); if (c->record_fields + r->fields > RECORD_FIELDS) - fail_at(c, r->used_line, r->used_column, "at most 128 record fields are allowed (%s.%s has %u more)", library, + fail_at(c, r->used_line, r->used_column, "at most %d record fields are allowed (%s.%s has %u more)", RECORD_FIELDS, library, r->name, r->fields); c->records[c->record_count++] = *r; c->record_fields += r->fields; @@ -5258,7 +5266,7 @@ static void flatten(Compiler *c, unsigned index) { if (!r->variant || r->declared[i].kind || *r->declared[i].type) add_flat(c, r, r->declared[i].name, r->declared[i].kind ? r->declared[i].kind : NUM, r->declared[i].type, r->declared[i].number, r->declared[i].integer, r->declared[i].string, r->declared[i].line, r->declared[i].column); - if (c->record_fields + r->fields > RECORD_FIELDS) fail_at(c, r->line, r->column, "at most 128 record fields are allowed"); + if (c->record_fields + r->fields > RECORD_FIELDS) fail_at(c, r->line, r->column, "at most %d record fields are allowed", RECORD_FIELDS); c->record_fields += r->fields; r->flattening = false; r->flattened = true; @@ -5276,8 +5284,8 @@ static int scalar_record(Compiler *c, const char *name, int line, int column) { if (strcmp(name, scalars[k].name)) continue; int found = module_record(c, 0, name); if (found >= 0) return found; - if (c->record_count == RECORDS) fail_at(c, line, column, "at most 16 records are allowed, a list of %s among them", name); - if (c->record_fields == RECORD_FIELDS) fail_at(c, line, column, "at most 128 record fields are allowed"); + if (c->record_count == RECORDS) fail_at(c, line, column, "at most %d records are allowed, a list of %s among them", RECORDS, name); + if (c->record_fields == RECORD_FIELDS) fail_at(c, line, column, "at most %d record fields are allowed", RECORD_FIELDS); Record *r = &c->records[c->record_count]; memset(r, 0, sizeof(*r)); memcpy(r->name, name, strlen(name) + 1); @@ -5353,7 +5361,7 @@ static void generate(Compiler *c, unsigned char **module, size_t *size) { char name[NAME]; snprintf(name, sizeof(name), release ? "%s.up" : "frame", c->handlers[i].input); if (input_index(c, name) < 0) { - if (c->input_count == INPUTS) fail_at(c, c->handlers[i].line, c->handlers[i].column, "at most 32 inputs are allowed"); + if (c->input_count == INPUTS) fail_at(c, c->handlers[i].line, c->handlers[i].column, "at most %d inputs are allowed", INPUTS); Input *in = &c->inputs[c->input_count++]; memset(in, 0, sizeof(*in)); memcpy(in->name, name, NAME); @@ -5378,8 +5386,8 @@ static void generate(Compiler *c, unsigned char **module, size_t *size) { table = (int)t; } if (anchor < 0 || c->state_types[anchor] != STATE_INDEX || table < 0) continue; - if (c->input_count == INPUTS) fail_at(c, c->view_line[v][0], c->view_column[v][0], "at most 32 inputs are allowed"); - if (c->handler_count == HANDLERS) fail_at(c, c->view_line[v][0], c->view_column[v][0], "at most 64 handlers are allowed"); + if (c->input_count == INPUTS) fail_at(c, c->view_line[v][0], c->view_column[v][0], "at most %d inputs are allowed", INPUTS); + if (c->handler_count == HANDLERS) fail_at(c, c->view_line[v][0], c->view_column[v][0], "at most %d handlers are allowed", HANDLERS); Input *in = &c->inputs[c->input_count++]; memset(in, 0, sizeof(*in)); char point[NAME + 8]; @@ -5692,7 +5700,6 @@ bool aup_compile(const char *source, size_t length, AupLoad load, void *context, c->module_count = 1; /* the main source */ unsigned char *result = NULL; if (!setjmp(c->fail)) { - if (length > AUP_SOURCE_LIMIT) fail_at(c, 1, 1, "source exceeds 32768 bytes"); c->token = scan(c); separators(c); while (c->token.kind != T_EOF) declaration(c); diff --git a/compiler/main.c b/compiler/main.c index b822430..212dda4 100644 --- a/compiler/main.c +++ b/compiler/main.c @@ -14,16 +14,28 @@ static int usage(void) { return 2; } -/* Read a whole source (up to one byte past the limit, which the compiler - * reports); NULL if it cannot be read. */ +/* All of a file's bytes, of any size (Pure 027's 4f-4); NULL if they cannot be read. */ +static char *read_all(FILE *file, size_t *length) { + size_t room = 1 << 15, used = 0; + char *bytes = malloc(room); + while (bytes) { + used += fread(bytes + used, 1, room - used, file); + if (used < room || ferror(file)) break; + char *grown = room <= SIZE_MAX / 2 ? realloc(bytes, room *= 2) : NULL; + if (!grown) { free(bytes); bytes = NULL; } + else bytes = grown; + } + if (bytes && ferror(file)) { free(bytes); bytes = NULL; } + *length = bytes ? used : 0; + return bytes; +} + +/* Read a whole source; NULL if it cannot be read. */ static char *read_source(const char *path, size_t *length) { FILE *file = fopen(path, "rb"); if (!file) return NULL; - char *source = malloc(AUP_SOURCE_LIMIT + 1); - *length = source ? fread(source, 1, AUP_SOURCE_LIMIT + 1, file) : 0; - bool failed = !source || ferror(file); + char *source = read_all(file, length); fclose(file); - if (failed) { free(source); return NULL; } return source; } @@ -137,8 +149,7 @@ static int query(const char *app, const char *query_path, const char *out) { if (!app_source) { perror(app); return 2; } char *query_source = NULL; if (strcmp(query_path, "-")) query_source = read_input(&inputs, query_path, &query_length); - else if ((query_source = malloc(AUP_SOURCE_LIMIT + 1))) - query_length = fread(query_source, 1, AUP_SOURCE_LIMIT + 1, stdin); + else query_source = read_all(stdin, &query_length); char *source = query_source ? malloc(app_length + 1 + query_length) : NULL; if (!source) { perror(query_path); diff --git a/tests/check.py b/tests/check.py index 05977e2..7d53e1d 100644 --- a/tests/check.py +++ b/tests/check.py @@ -185,7 +185,7 @@ DIAGNOSTICS = [ ('\n'.join(['fn f0(x) = x + 1'] + [f'fn f{i}(x) = f{i - 1}(f{i - 1}(x))' for i in range(1, 15)] + ['view v = f14(1)']) + '\n', '1:14', "program exceeds 65535 instructions (in fn 'f0' called at 2:15)"), ('view v = ' + '(' * 70 + '1' + ')' * 70 + '\n', '1:', 'expression nesting exceeds 64'), - ('state s = 0\n' * 1 + ''.join(f'state s{i} = 0\n' for i in range(32)), '33:7', 'at most 32 states are allowed'), + ('state s = 0\n' * 1 + ''.join(f'state s{i} = 0\n' for i in range(4096)), '4097:7', 'at most 4096 states are allowed'), # v0b: messages and types ('input m = message { a: u32 }\non m {\n reply m.a\n}\n', '3:9', 'reply takes bytes or text, not a number'), ('input m = message {\n a: float\n}\n', '2:6', 'field types are u32, i64, number, bool, bytes, text and list of a record'), @@ -2416,6 +2416,20 @@ def main(): report['diagnostics'] = len(DIAGNOSTICS) report['accepted_forms'] = len(ACCEPTED) + # No fixed counts but Pure's widths (Pure 027's 4f): 300 states, 40 inputs, 70 views and 20 records, in a source past + # 32 KB (40 KB of comments), compiled and run to the last of each. + wide = ''.join(f'record R{r} {{\n a = 0\n}}\n' for r in range(20)) + 'state items: list of R19\n' + wide += ''.join(f'state s{i} = 0\n' for i in range(300)) + ''.join(f'input m{i} = message {{ n: u32 }}\n' for i in range(40)) + wide += 'on m39 {\n s299 = s299 + m39.n\n items.append(R19 { a: m39.n })\n}\n' + wide += ''.join(f'view v{i} = s{i}\n' for i in range(69)) + 'view v69 = items.count\n' + wide += ''.join(f'# {i:05} ' + 'x' * 60 + '\n' for i in range(600)) + assert len(wide) > 40000 + (work / 'wide.aup').write_text(wide) + compile_module(aup, work / 'wide.aup', work / 'wide.pmod') + shown = json.loads(run(pure_run, work / 'wide.pmod', stdin='message m39 05000000\nmessage m39 02000000\n').stdout.splitlines()[-1]) + assert shown['error'] is None and shown['states']['s299'] == 7 and shown['views']['v69'] == 2, shown['states']['s299'] + report['wide_program'] = {'states': 301, 'inputs': 40, 'views': 70, 'records': 20, 'source_bytes': len(wide)} + # message like X is X's fields, written once: the same module, byte for byte. fields = ' title: text\n lane: u32\n to: {\n city: text\n }\n' body = 'on remote {\n n = remote.lane\n}\non tap.down {\n send changed { title: "t", lane: 1, to: { city: "c" } }\n}\n' |