micropython · dpgeorge · Jul 9, 2025 · Jun 26, 2025 · Jun 26, 2025 · Mar 21, 2025
diff --git a/py/bc.h b/py/bc.h
@@ -308,25 +308,35 @@ static inline void mp_module_context_alloc_tables(mp_module_context_t *context,
     #endif
 }
 
+typedef struct _mp_code_lineinfo_t {
+    size_t bc_increment;
+    size_t line_increment;
+} mp_code_lineinfo_t;
+
+static inline mp_code_lineinfo_t mp_bytecode_decode_lineinfo(const byte **line_info) {
+    mp_code_lineinfo_t result;
+    size_t c = (*line_info)[0];
+    if ((c & 0x80) == 0) {
+        // 0b0LLBBBBB encoding
+        result.bc_increment = c & 0x1f;
+        result.line_increment = c >> 5;
+        *line_info += 1;
+    } else {
+        // 0b1LLLBBBB 0bLLLLLLLL encoding (l's LSB in second byte)
+        result.bc_increment = c & 0xf;
+        result.line_increment = ((c << 4) & 0x700) | (*line_info)[1];
+        *line_info += 2;
+    }
+    return result;
+}
+
 static inline size_t mp_bytecode_get_source_line(const byte *line_info, const byte *line_info_top, size_t bc_offset) {
     size_t source_line = 1;
     while (line_info < line_info_top) {
-        size_t c = *line_info;
-        size_t b, l;
-        if ((c & 0x80) == 0) {
-            // 0b0LLBBBBB encoding
-            b = c & 0x1f;
-            l = c >> 5;
-            line_info += 1;
-        } else {
-            // 0b1LLLBBBB 0bLLLLLLLL encoding (l's LSB in second byte)
-            b = c & 0xf;
-            l = ((c << 4) & 0x700) | line_info[1];
-            line_info += 2;
-        }
-        if (bc_offset >= b) {
-            bc_offset -= b;
-            source_line += l;
+        mp_code_lineinfo_t decoded = mp_bytecode_decode_lineinfo(&line_info);
+        if (bc_offset >= decoded.bc_increment) {
+            bc_offset -= decoded.bc_increment;
+            source_line += decoded.line_increment;
         } else {
             // found source line corresponding to bytecode offset
             break;

diff --git a/py/objcode.c b/py/objcode.c
@@ -107,6 +107,67 @@ static mp_obj_t raw_code_lnotab(const mp_raw_code_t *rc) {
     return o;
 }
 
+static mp_obj_t code_colines_iter(mp_obj_t);
+static mp_obj_t code_colines_next(mp_obj_t);
+typedef struct _mp_obj_colines_iter_t {
+    mp_obj_base_t base;
+    mp_fun_1_t iternext;
+    const mp_raw_code_t *rc;
+    mp_uint_t bc;
+    mp_uint_t source_line;
+    const byte *ci;
+} mp_obj_colines_iter_t;
+
+static mp_obj_t code_colines_iter(mp_obj_t self_in) {
+    mp_obj_code_t *self = MP_OBJ_TO_PTR(self_in);
+    mp_obj_colines_iter_t *iter = mp_obj_malloc(mp_obj_colines_iter_t, &mp_type_polymorph_iter);
+    iter->iternext = code_colines_next;
+    iter->rc = self->rc;
+    iter->bc = 0;
+    iter->source_line = 1;
+    iter->ci = self->rc->prelude.line_info;
+    return MP_OBJ_FROM_PTR(iter);
+}
+static MP_DEFINE_CONST_FUN_OBJ_1(code_colines_obj, code_colines_iter);
+
+static mp_obj_t code_colines_next(mp_obj_t iter_in) {
+    mp_obj_colines_iter_t *iter = MP_OBJ_TO_PTR(iter_in);
+    const byte *ci_end = iter->rc->prelude.line_info_top;
+
+    mp_uint_t start = iter->bc;
+    mp_uint_t line_no = iter->source_line;
+    bool another = true;
+
+    while (another && iter->ci < ci_end) {
+        another = false;
+        mp_code_lineinfo_t decoded = mp_bytecode_decode_lineinfo(&iter->ci);
+        iter->bc += decoded.bc_increment;
+        iter->source_line += decoded.line_increment;
+
+        if (decoded.bc_increment == 0) {
+            line_no = iter->source_line;
+            another = true;
+        } else if (decoded.line_increment == 0) {
+            another = true;
+        }
+    }
+
+    if (another) {
+        mp_uint_t prelude_size = (iter->rc->prelude.opcodes - (const byte *)iter->rc->fun_data);
+        mp_uint_t bc_end = iter->rc->fun_data_len - prelude_size;
+        if (iter->bc >= bc_end) {
+            return MP_OBJ_STOP_ITERATION;
+        } else {
+            iter->bc = bc_end;
+        }
+    }
+
+    mp_uint_t end = iter->bc;
+    mp_obj_t next[3] = {MP_OBJ_NEW_SMALL_INT(start), MP_OBJ_NEW_SMALL_INT(end), MP_OBJ_NEW_SMALL_INT(line_no)};
+
+    return mp_obj_new_tuple(MP_ARRAY_SIZE(next), next);
+}
+
 static void code_attr(mp_obj_t self_in, qstr attr, mp_obj_t *dest) {
     if (dest[0] != MP_OBJ_NULL) {
         // not load attribute
@@ -143,6 +204,10 @@ static void code_attr(mp_obj_t self_in, qstr attr, mp_obj_t *dest) {
             }
             dest[0] = o->lnotab;
             break;
+        case MP_QSTR_co_lines:
+            dest[0] = MP_OBJ_FROM_PTR(&code_colines_obj);
+            dest[1] = self_in;
+            break;
     }
 }
 

diff --git a/py/showbc.c b/py/showbc.c
@@ -144,17 +144,9 @@ void mp_bytecode_print(const mp_print_t *print, const mp_raw_code_t *rc, size_t
         mp_uint_t source_line = 1;
         mp_printf(print, "  bc=" INT_FMT " line=" UINT_FMT "\n", bc, source_line);
         for (const byte *ci = code_info; ci < line_info_top;) {
-            if ((ci[0] & 0x80) == 0) {
-                // 0b0LLBBBBB encoding
-                bc += ci[0] & 0x1f;
-                source_line += ci[0] >> 5;
-                ci += 1;
-            } else {
-                // 0b1LLLBBBB 0bLLLLLLLL encoding (l's LSB in second byte)
-                bc += ci[0] & 0xf;
-                source_line += ((ci[0] << 4) & 0x700) | ci[1];
-                ci += 2;
-            }
+            mp_code_lineinfo_t decoded = mp_bytecode_decode_lineinfo(&ci);
+            bc += decoded.bc_increment;
+            source_line += decoded.line_increment;
             mp_printf(print, "  bc=" INT_FMT " line=" UINT_FMT "\n", bc, source_line);
         }
     }

diff --git a/tests/basics/fun_code_colines.py b/tests/basics/fun_code_colines.py
@@ -0,0 +1,81 @@
+# Check that we have sensical bytecode offsets in function.__code__.co_lines
+
+def f1(x, y, obj, obj2, obj3):
+    a = x + y # line 4: bc+4 line+4
+    b = x - y # line 5: bc+4 line+1
+    # line 6
+    # line 7
+    # line 8
+    # line 9
+    # line 10
+    # line 11
+    # line 12
+    # line 13
+    # line 14
+    # line 15
+    # line 16
+    # line 17
+    # line 18
+    # line 19
+    c = a * b # line 20: bc+4 line+15
+    obj.a.b.c.d.e.f.g.h.i.j.k.l.m.n.o.p.q.r.s.t.u.v.w.x.y.z.fun() # line 21: bc+31 line+1; bc+27 line+0
+    # line 22
+    # line 23
+    # line 24: bc+0 line+3
+    # line 25
+    # line 26
+    # line 27: bc+0 line+3
+    # line 28
+    # line 29
+    obj2.a.b.c.d.e.f.g.h.i.j.k.l.m.n.o.p.q.r.s.t.u.v.w.x.y.z.fun() # line 30: bc+31 line+3; bc+27 line+0
+    # line 31
+    # line 32
+    # line 33: bc+0 line+3
+    # line 34
+    # line 35
+    # line 36
+    # line 37
+    # line 38
+    # line 39
+    # line 40
+    # line 41
+    # line 42
+    # line 43
+    # line 44
+    # line 45
+    # line 46
+    # line 47
+    # line 48
+    # line 49
+    # line 50
+    # line 51
+    # line 52
+    # line 53
+    # line 54
+    # line 55
+    # line 56
+    # line 57
+    # line 58
+    # line 59
+    return obj3.a.b.c.d.e.f.g.h.i.j.k.l.m.n.o.p.q.r.s.t.u.v.w.x.y.z.fun() # line 60: bc+31 line+27; bc+27 line+0
+
+def f2(x, y):
+    a = x + y # line 63
+    b = x - y # line 64
+    return a * b # line 65
+
+try:
+    f1.__code__.co_lines
+except AttributeError:
+    print("SKIP")
+    raise SystemExit
+
+print("f1")
+for start, end, line_no in f1.__code__.co_lines():
+    print("line {} start: {}".format(line_no, start))
+    print("line {} end: {}".format(line_no, end))
+
+print("f2")
+for start, end, line_no in f2.__code__.co_lines():
+    print("line {} start: {}".format(line_no, start))
+    print("line {} end: {}".format(line_no, end))
diff --git a/tests/basics/fun_code_colines.py.exp b/tests/basics/fun_code_colines.py.exp
@@ -0,0 +1,20 @@
+f1
+line 4 start: 0
+line 4 end: 4
+line 5 start: 4
+line 5 end: 8
+line 20 start: 8
+line 20 end: 12
+line 21 start: 12
+line 21 end: 70
+line 30 start: 70
+line 30 end: 128
+line 60 start: 128
+line 60 end: 186
+f2
+line 63 start: 0
+line 63 end: 4
+line 64 start: 4
+line 64 end: 8
+line 65 start: 8
+line 65 end: 12
diff --git a/tests/basics/fun_code_full.py b/tests/basics/fun_code_full.py
@@ -0,0 +1,47 @@
+# Test function.__code__ attributes not available with MICROPY_PY_BUILTINS_CODE <= MICROPY_PY_BUILTINS_CODE_BASIC
+
+try:
+    (lambda: 0).__code__.co_code
+except AttributeError:
+    print("SKIP")
+    raise SystemExit
+
+try:
+    import warnings
+    warnings.simplefilter("ignore") # ignore deprecation warning about co_lnotab
+except ImportError:
+    pass
+
+def f(x, y):
+    a = x + y
+    b = x - y
+    return a * b
+
+code = f.__code__
+
+print(type(code.co_code)) # both bytes (but mpy and cpy have different instruction sets)
+print(code.co_consts) # (not necessarily the same set, but in this function they are)
+print(code.co_filename.rsplit('/')[-1]) # same terminal filename but might be different paths on other ports
+print(type(code.co_firstlineno)) # both ints (but mpy points to first line inside, cpy points to declaration)
+print(code.co_name)
+print(iter(code.co_names) is not None) # both iterable (but mpy returns dict with names as keys, cpy only the names; and not necessarily the same set)
+print(type(code.co_lnotab)) # both bytes
+
+co_lines = code.co_lines()
+
+l = list(co_lines)
+first_start = l[0][0]
+last_end = l[-1][1]
+print(first_start) # co_lines should start at the start of the bytecode
+print(len(code.co_code) - last_end) # and end at the end of the bytecode
+
+prev_end = 0
+for start, end, line_no in l:
+    if end != prev_end:
+        print("non-contiguous")
+        break # the offset ranges should be contiguous
+    prev_end = end
+else:
+    print("contiguous")
+
+