djsundog/circuitpython
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
4e75aaecd0
circuitpython/py/objmodule.c

290 lines
10 KiB
C
Raw Normal View History

Add license header to (almost) all files. Blanket wide to all .c and .h files. Some files originating from ST are difficult to deal with (license wise) so it was left out of those. Also merged modpyb.h, modos.h, modstm.h and modtime.h in stmhal/.
2014-05-03 18:27:38 -04:00
/*
all: Use the name MicroPython consistently in comments There were several different spellings of MicroPython present in comments, when there should be only one.
2017-06-30 03:22:17 -04:00
* This file is part of the MicroPython project, http://micropython.org/
Add license header to (almost) all files. Blanket wide to all .c and .h files. Some files originating from ST are difficult to deal with (license wise) so it was left out of those. Also merged modpyb.h, modos.h, modstm.h and modtime.h in stmhal/.
2014-05-03 18:27:38 -04:00
*
* The MIT License (MIT)
*
* Copyright (c) 2013, 2014 Damien P. George
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
Add module object, to be used eventually for import.
2014-01-02 16:30:26 -05:00
#include <stdlib.h>
#include <assert.h>
Introduce a long lived section of the heap. This adapts the allocation process to start from either end of the heap when searching for free space. The default behavior is identical to the existing behavior where it starts with the lowest block and looks higher. Now it can also look from the highest block and lower depending on the long_lived parameter to gc_alloc. As the heap fills, the two sections may overlap. When they overlap, a collect may be triggered in order to keep the long lived section compact. However, free space is always eligable for each type of allocation. By starting from either of the end of the heap we have ability to separate short lived objects from long lived ones. This separation reduces heap fragmentation because long lived objects are easy to densely pack. Most objects are short lived initially but may be made long lived when they are referenced by a type or module. This involves copying the memory and then letting the collect phase free the old portion. QSTR pools and chunks are always long lived because they are never freed. The reallocation, collection and free processes are largely unchanged. They simply also maintain an index to the highest free block as well as the lowest. These indices are used to speed up the allocation search until the next collect. In practice, this change may slightly slow down import statements with the benefit that memory is much less fragmented afterwards. For example, a test import into a 20k heap that leaves ~6k free previously had the largest continuous free space of ~400 bytes. After this change, the largest continuous free space is over 3400 bytes.
2018-01-23 19:22:05 -05:00
#include "py/gc.h"
py: Move to guarded includes, everywhere in py/ core. Addresses issue #1022.
2015-01-01 15:27:54 -05:00
#include "py/objmodule.h"
#include "py/runtime.h"
#include "py/builtin.h"
Proper support for registering builtin modules in ROM. Comes with some refactoring of code and renaming of files. All modules are now named mod*.[ch].
2014-03-25 10:18:18 -04:00
py: Overhaul and simplify printf/pfenv mechanism. Previous to this patch the printing mechanism was a bit of a tangled mess. This patch attempts to consolidate printing into one interface. All (non-debug) printing now uses the mp_print* family of functions, mainly mp_printf. All these functions take an mp_print_t structure as their first argument, and this structure defines the printing backend through the "print_strn" function of said structure. Printing from the uPy core can reach the platform-defined print code via two paths: either through mp_sys_stdout_obj (defined pert port) in conjunction with mp_stream_write; or through the mp_plat_print structure which uses the MP_PLAT_PRINT_STRN macro to define how string are printed on the platform. The former is only used when MICROPY_PY_IO is defined. With this new scheme printing is generally more efficient (less layers to go through, less arguments to pass), and, given an mp_print_t* structure, one can call mp_print_str for efficiency instead of mp_printf("%s", ...). Code size is also reduced by around 200 bytes on Thumb2 archs.
2015-04-09 18:56:15 -04:00
STATIC void module_print(const mp_print_t *print, mp_obj_t self_in, mp_print_kind_t kind) {
py, unix: Allow to compile with -Wunused-parameter. See issue #699.
2015-01-20 07:47:20 -05:00
(void)kind;
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
mp_obj_module_t *self = MP_OBJ_TO_PTR(self_in);
py: Print imported module's location (__file__) if available.
2014-09-08 05:45:23 -04:00
all: Remove 'name' member from mp_obj_module_t struct. One can instead lookup __name__ in the modules dict to get the value.
2016-09-20 20:52:53 -04:00
const char *module_name = "";
mp_map_elem_t *elem = mp_map_lookup(&self->globals->map, MP_OBJ_NEW_QSTR(MP_QSTR___name__), MP_MAP_LOOKUP);
if (elem != NULL) {
module_name = mp_obj_str_get_str(elem->value);
}
py: Print imported module's location (__file__) if available.
2014-09-08 05:45:23 -04:00
#if MICROPY_PY___FILE__
// If we store __file__ to imported modules then try to lookup this
// symbol to give more information about the module.
all: Remove 'name' member from mp_obj_module_t struct. One can instead lookup __name__ in the modules dict to get the value.
2016-09-20 20:52:53 -04:00
elem = mp_map_lookup(&self->globals->map, MP_OBJ_NEW_QSTR(MP_QSTR___file__), MP_MAP_LOOKUP);
py: Print imported module's location (__file__) if available.
2014-09-08 05:45:23 -04:00
if (elem != NULL) {
all: Remove 'name' member from mp_obj_module_t struct. One can instead lookup __name__ in the modules dict to get the value.
2016-09-20 20:52:53 -04:00
mp_printf(print, "<module '%s' from '%s'>", module_name, mp_obj_str_get_str(elem->value));
py: Print imported module's location (__file__) if available.
2014-09-08 05:45:23 -04:00
return;
}
#endif
all: Remove 'name' member from mp_obj_module_t struct. One can instead lookup __name__ in the modules dict to get the value.
2016-09-20 20:52:53 -04:00
mp_printf(print, "<module '%s'>", module_name);
Add module object, to be used eventually for import.
2014-01-02 16:30:26 -05:00
}
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
STATIC void module_attr(mp_obj_t self_in, qstr attr, mp_obj_t *dest) {
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
mp_obj_module_t *self = MP_OBJ_TO_PTR(self_in);
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
if (dest[0] == MP_OBJ_NULL) {
// load attribute
mp_map_elem_t *elem = mp_map_lookup(&self->globals->map, MP_OBJ_NEW_QSTR(attr), MP_MAP_LOOKUP);
if (elem != NULL) {
dest[0] = elem->value;
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
}
py: Finish implementation of all del opcodes. At this point, all opcodes are now implemented! Some del opcodes have been combined with store opcodes, with the value to store being MP_OBJ_NULL.
2014-04-08 16:11:49 -04:00
} else {
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
// delete/store attribute
mp_obj_dict_t *dict = self->globals;
if (dict->map.is_fixed) {
#if MICROPY_CAN_OVERRIDE_BUILTINS
if (dict == &mp_module_builtins_globals) {
if (MP_STATE_VM(mp_module_builtins_override_dict) == NULL) {
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
MP_STATE_VM(mp_module_builtins_override_dict) = MP_OBJ_TO_PTR(mp_obj_new_dict(1));
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
}
dict = MP_STATE_VM(mp_module_builtins_override_dict);
} else
#endif
{
// can't delete or store to fixed map
return;
}
}
if (dest[1] == MP_OBJ_NULL) {
// delete attribute
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
mp_obj_dict_delete(MP_OBJ_FROM_PTR(dict), MP_OBJ_NEW_QSTR(attr));
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
} else {
// store attribute
Introduce a long lived section of the heap. This adapts the allocation process to start from either end of the heap when searching for free space. The default behavior is identical to the existing behavior where it starts with the lowest block and looks higher. Now it can also look from the highest block and lower depending on the long_lived parameter to gc_alloc. As the heap fills, the two sections may overlap. When they overlap, a collect may be triggered in order to keep the long lived section compact. However, free space is always eligable for each type of allocation. By starting from either of the end of the heap we have ability to separate short lived objects from long lived ones. This separation reduces heap fragmentation because long lived objects are easy to densely pack. Most objects are short lived initially but may be made long lived when they are referenced by a type or module. This involves copying the memory and then letting the collect phase free the old portion. QSTR pools and chunks are always long lived because they are never freed. The reallocation, collection and free processes are largely unchanged. They simply also maintain an index to the highest free block as well as the lowest. These indices are used to speed up the allocation search until the next collect. In practice, this change may slightly slow down import statements with the benefit that memory is much less fragmented afterwards. For example, a test import into a 20k heap that leaves ~6k free previously had the largest continuous free space of ~400 bytes. After this change, the largest continuous free space is over 3400 bytes.
2018-01-23 19:22:05 -05:00
mp_obj_t long_lived = gc_make_long_lived(dest[1]);
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
// TODO CPython allows STORE_ATTR to a module, but is this the correct implementation?
Introduce a long lived section of the heap. This adapts the allocation process to start from either end of the heap when searching for free space. The default behavior is identical to the existing behavior where it starts with the lowest block and looks higher. Now it can also look from the highest block and lower depending on the long_lived parameter to gc_alloc. As the heap fills, the two sections may overlap. When they overlap, a collect may be triggered in order to keep the long lived section compact. However, free space is always eligable for each type of allocation. By starting from either of the end of the heap we have ability to separate short lived objects from long lived ones. This separation reduces heap fragmentation because long lived objects are easy to densely pack. Most objects are short lived initially but may be made long lived when they are referenced by a type or module. This involves copying the memory and then letting the collect phase free the old portion. QSTR pools and chunks are always long lived because they are never freed. The reallocation, collection and free processes are largely unchanged. They simply also maintain an index to the highest free block as well as the lowest. These indices are used to speed up the allocation search until the next collect. In practice, this change may slightly slow down import statements with the benefit that memory is much less fragmented afterwards. For example, a test import into a 20k heap that leaves ~6k free previously had the largest continuous free space of ~400 bytes. After this change, the largest continuous free space is over 3400 bytes.
2018-01-23 19:22:05 -05:00
mp_obj_dict_store(MP_OBJ_FROM_PTR(dict), MP_OBJ_NEW_QSTR(attr), long_lived);
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
}
dest[0] = MP_OBJ_NULL; // indicate success
py: Finish implementation of all del opcodes. At this point, all opcodes are now implemented! Some del opcodes have been combined with store opcodes, with the value to store being MP_OBJ_NULL.
2014-04-08 16:11:49 -04:00
}
Improved type/class/instance code; mp_obj_type_t now has load_attr, store_attr. Creating of classes (types) and instances is much more like CPython now. You can use "type('name', (), {...})" to create classes.
2014-01-09 15:57:50 -05:00
}
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
const mp_obj_type_t mp_type_module = {
Implement proper exception type hierarchy. Each built-in exception is now a type, with base type BaseException. C exceptions are created by passing a pointer to the exception type to make an instance of. When raising an exception from the VM, an instance is created automatically if an exception type is raised (as opposed to an exception instance). Exception matching (RT_BINARY_OP_EXCEPTION_MATCH) is now proper. Handling of parse error changed to match new exceptions. mp_const_type renamed to mp_type_type for consistency.
2014-02-15 11:10:44 -05:00
{ &mp_type_type },
Change mp_obj_type_t.name from const char * to qstr. Ultimately all static strings should be qstr. This entry in the type structure is only used for printing error messages (to tell the type of the bad argument), and printing objects that don't supply a .print method.
2014-02-15 06:34:50 -05:00
.name = MP_QSTR_module,
Merge branch 'cplusplus' of https://github.com/ian-v/micropython into ian-v-cplusplus Conflicts: py/objcomplex.c
2014-01-07 10:58:30 -05:00
.print = module_print,
py: Combine load_attr and store_attr type methods into one (attr). This simplifies the API for objects and reduces code size (by around 400 bytes on Thumb2, and around 2k on x86). Performance impact was measured with Pystone score, but change was barely noticeable.
2015-04-01 10:10:50 -04:00
.attr = module_attr,
Add module object, to be used eventually for import.
2014-01-02 16:30:26 -05:00
};
mp_obj_t mp_obj_new_module(qstr module_name) {
py/modsys: Implement sys.modules. This for example will allow people to reload modules which didn't load successfully (e.g. due to syntax error).
2015-12-04 17:09:10 -05:00
mp_map_t *mp_loaded_modules_map = &MP_STATE_VM(mp_loaded_modules_dict).map;
mp_map_elem_t *el = mp_map_lookup(mp_loaded_modules_map, MP_OBJ_NEW_QSTR(module_name), MP_MAP_LOOKUP_ADD_IF_NOT_FOUND);
Implement modules as singletons Python semantics. In Python, importing module several times returns same underlying module object. This also fixes import statement handling for builtin modules. There're still issues: 1. CPython exposes set of loaded modules as sys.modules, we may want to do that either. 2. Builtin modules are implicitly imported, which is not really correct. We should separate registering a (builtin) module and importing a module. CPython keeps builtin module names in sys.builtin_module_names .
2014-01-19 17:03:34 -05:00
// We could error out if module already exists, but let C extensions
// add new members to existing modules.
if (el->value != MP_OBJ_NULL) {
return el->value;
}
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
// create new module object
Introduce a long lived section of the heap. This adapts the allocation process to start from either end of the heap when searching for free space. The default behavior is identical to the existing behavior where it starts with the lowest block and looks higher. Now it can also look from the highest block and lower depending on the long_lived parameter to gc_alloc. As the heap fills, the two sections may overlap. When they overlap, a collect may be triggered in order to keep the long lived section compact. However, free space is always eligable for each type of allocation. By starting from either of the end of the heap we have ability to separate short lived objects from long lived ones. This separation reduces heap fragmentation because long lived objects are easy to densely pack. Most objects are short lived initially but may be made long lived when they are referenced by a type or module. This involves copying the memory and then letting the collect phase free the old portion. QSTR pools and chunks are always long lived because they are never freed. The reallocation, collection and free processes are largely unchanged. They simply also maintain an index to the highest free block as well as the lowest. These indices are used to speed up the allocation search until the next collect. In practice, this change may slightly slow down import statements with the benefit that memory is much less fragmented afterwards. For example, a test import into a 20k heap that leaves ~6k free previously had the largest continuous free space of ~400 bytes. After this change, the largest continuous free space is over 3400 bytes.
2018-01-23 19:22:05 -05:00
mp_obj_module_t *o = m_new_ll_obj(mp_obj_module_t);
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
o->base.type = &mp_type_module;
Introduce a long lived section of the heap. This adapts the allocation process to start from either end of the heap when searching for free space. The default behavior is identical to the existing behavior where it starts with the lowest block and looks higher. Now it can also look from the highest block and lower depending on the long_lived parameter to gc_alloc. As the heap fills, the two sections may overlap. When they overlap, a collect may be triggered in order to keep the long lived section compact. However, free space is always eligable for each type of allocation. By starting from either of the end of the heap we have ability to separate short lived objects from long lived ones. This separation reduces heap fragmentation because long lived objects are easy to densely pack. Most objects are short lived initially but may be made long lived when they are referenced by a type or module. This involves copying the memory and then letting the collect phase free the old portion. QSTR pools and chunks are always long lived because they are never freed. The reallocation, collection and free processes are largely unchanged. They simply also maintain an index to the highest free block as well as the lowest. These indices are used to speed up the allocation search until the next collect. In practice, this change may slightly slow down import statements with the benefit that memory is much less fragmented afterwards. For example, a test import into a 20k heap that leaves ~6k free previously had the largest continuous free space of ~400 bytes. After this change, the largest continuous free space is over 3400 bytes.
2018-01-23 19:22:05 -05:00
o->globals = MP_OBJ_TO_PTR(gc_make_long_lived(mp_obj_new_dict(MICROPY_MODULE_DICT_SIZE)));
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
// store __name__ entry in the module
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
mp_obj_dict_store(MP_OBJ_FROM_PTR(o->globals), MP_OBJ_NEW_QSTR(MP_QSTR___name__), MP_OBJ_NEW_QSTR(module_name));
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
// store the new module into the slot in the global dict holding all modules
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
el->value = MP_OBJ_FROM_PTR(o);
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
// return the new module
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
return MP_OBJ_FROM_PTR(o);
Add module object, to be used eventually for import.
2014-01-02 16:30:26 -05:00
}
py: Change module globals from mp_map_t* to mp_obj_dict_t*. Towards addressing issue #424. Had a small increase to ROM usage (order 60 bytes).
2014-04-05 16:53:54 -04:00
mp_obj_dict_t *mp_obj_module_get_globals(mp_obj_t self_in) {
Proper support for registering builtin modules in ROM. Comes with some refactoring of code and renaming of files. All modules are now named mod*.[ch].
2014-03-25 10:18:18 -04:00
assert(MP_OBJ_IS_TYPE(self_in, &mp_type_module));
py: Wrap all obj-ptr conversions in MP_OBJ_TO_PTR/MP_OBJ_FROM_PTR. This allows the mp_obj_t type to be configured to something other than a pointer-sized primitive type. This patch also includes additional changes to allow the code to compile when sizeof(mp_uint_t) != sizeof(void*), such as using size_t instead of mp_uint_t, and various casts.
2015-11-27 12:01:44 -05:00
mp_obj_module_t *self = MP_OBJ_TO_PTR(self_in);
Proper support for registering builtin modules in ROM. Comes with some refactoring of code and renaming of files. All modules are now named mod*.[ch].
2014-03-25 10:18:18 -04:00
return self->globals;
}
Introduce a long lived section of the heap. This adapts the allocation process to start from either end of the heap when searching for free space. The default behavior is identical to the existing behavior where it starts with the lowest block and looks higher. Now it can also look from the highest block and lower depending on the long_lived parameter to gc_alloc. As the heap fills, the two sections may overlap. When they overlap, a collect may be triggered in order to keep the long lived section compact. However, free space is always eligable for each type of allocation. By starting from either of the end of the heap we have ability to separate short lived objects from long lived ones. This separation reduces heap fragmentation because long lived objects are easy to densely pack. Most objects are short lived initially but may be made long lived when they are referenced by a type or module. This involves copying the memory and then letting the collect phase free the old portion. QSTR pools and chunks are always long lived because they are never freed. The reallocation, collection and free processes are largely unchanged. They simply also maintain an index to the highest free block as well as the lowest. These indices are used to speed up the allocation search until the next collect. In practice, this change may slightly slow down import statements with the benefit that memory is much less fragmented afterwards. For example, a test import into a 20k heap that leaves ~6k free previously had the largest continuous free space of ~400 bytes. After this change, the largest continuous free space is over 3400 bytes.
2018-01-23 19:22:05 -05:00
void mp_obj_module_set_globals(mp_obj_t self_in, mp_obj_dict_t *globals) {
assert(MP_OBJ_IS_TYPE(self_in, &mp_type_module));
mp_obj_module_t *self = MP_OBJ_TO_PTR(self_in);
self->globals = globals;
}
Proper support for registering builtin modules in ROM. Comes with some refactoring of code and renaming of files. All modules are now named mod*.[ch].
2014-03-25 10:18:18 -04:00
/******************************************************************************/
// Global module table and related functions
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
STATIC const mp_rom_map_elem_t mp_builtin_module_table[] = {
{ MP_ROM_QSTR(MP_QSTR___main__), MP_ROM_PTR(&mp_module___main__) },
{ MP_ROM_QSTR(MP_QSTR_builtins), MP_ROM_PTR(&mp_module_builtins) },
{ MP_ROM_QSTR(MP_QSTR_micropython), MP_ROM_PTR(&mp_module_micropython) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#if MICROPY_PY_ARRAY
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_array), MP_ROM_PTR(&mp_module_array) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_IO
regularize how module weak links and alternate names are listed
2019-02-18 00:45:43 -05:00
{ MP_ROM_QSTR(MP_QSTR_io), MP_ROM_PTR(&mp_module_io) },
py/modio: Rename module name to "uio" for consistency with other modules.
2016-05-02 06:56:33 -04:00
{ MP_ROM_QSTR(MP_QSTR_uio), MP_ROM_PTR(&mp_module_io) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_COLLECTIONS
Updated library source: ucollections -> collections
2018-05-14 14:33:50 -04:00
{ MP_ROM_QSTR(MP_QSTR_collections), MP_ROM_PTR(&mp_module_collections) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_STRUCT
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_ustruct), MP_ROM_PTR(&mp_module_ustruct) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_BUILTINS_FLOAT
#if MICROPY_PY_MATH
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_math), MP_ROM_PTR(&mp_module_math) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
py: Make it easy to build without MICROPY_PY_BUILTINS_COMPLEX. Automagically skip related modules.
2015-12-06 17:18:44 -05:00
#if MICROPY_PY_BUILTINS_COMPLEX && MICROPY_PY_CMATH
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_cmath), MP_ROM_PTR(&mp_module_cmath) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#endif
#if MICROPY_PY_SYS
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_sys), MP_ROM_PTR(&mp_module_sys) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_GC && MICROPY_ENABLE_GC
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_gc), MP_ROM_PTR(&mp_module_gc) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
py: Add basic _thread module, with ability to start a new thread.
2016-04-22 18:52:33 -04:00
#if MICROPY_PY_THREAD
{ MP_ROM_QSTR(MP_QSTR__thread), MP_ROM_PTR(&mp_module_thread) },
#endif
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
// extmod modules
py: Add uerrno module, with errno constants and dict.
2016-05-10 05:54:25 -04:00
#if MICROPY_PY_UERRNO
{ MP_ROM_QSTR(MP_QSTR_uerrno), MP_ROM_PTR(&mp_module_uerrno) },
#endif
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#if MICROPY_PY_UCTYPES
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_uctypes), MP_ROM_PTR(&mp_module_uctypes) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_UZLIB
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_uzlib), MP_ROM_PTR(&mp_module_uzlib) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_UJSON
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_ujson), MP_ROM_PTR(&mp_module_ujson) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_URE
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_ure), MP_ROM_PTR(&mp_module_ure) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_UHEAPQ
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_uheapq), MP_ROM_PTR(&mp_module_uheapq) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
extmod/modutimeq: Refactor into optimized class. import utimeq, utime # Max queue size, the queue allocated statically on creation q = utimeq.utimeq(10) q.push(utime.ticks_ms(), data1, data2) res = [0, 0, 0] # Items in res are filled up with results q.pop(res)
2016-12-21 16:29:32 -05:00
#if MICROPY_PY_UTIMEQ
{ MP_ROM_QSTR(MP_QSTR_utimeq), MP_ROM_PTR(&mp_module_utimeq) },
#endif
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#if MICROPY_PY_UHASHLIB
uhashlib: masquerade as hashlib for python3 compatibility
2018-05-06 13:29:55 -04:00
{ MP_ROM_QSTR(MP_QSTR_hashlib), MP_ROM_PTR(&mp_module_uhashlib) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
#if MICROPY_PY_UBINASCII
ubinascii: masquerade as binascii for python3 compatibility
2018-04-30 21:03:38 -04:00
{ MP_ROM_QSTR(MP_QSTR_binascii), MP_ROM_PTR(&mp_module_ubinascii) },
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
#endif
extmod/modurandom: Add "urandom" module. Seedable and reproducible pseudo-random number generator. Implemented functions are getrandbits(n) (n <= 32) and seed(). The algorithm used is Yasmarang by Ilya Levin: http://www.literatecode.com/yasmarang
2016-01-17 05:10:28 -05:00
#if MICROPY_PY_URANDOM
{ MP_ROM_QSTR(MP_QSTR_urandom), MP_ROM_PTR(&mp_module_urandom) },
#endif
stmhal/moduselect: Move to extmod/ for reuse by other ports.
2016-11-20 15:49:45 -05:00
#if MICROPY_PY_USELECT
{ MP_ROM_QSTR(MP_QSTR_uselect), MP_ROM_PTR(&mp_module_uselect) },
#endif
modussl: SSL socket wrapper module based on axTLS.
2015-10-06 11:10:00 -04:00
#if MICROPY_PY_USSL
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_ussl), MP_ROM_PTR(&mp_module_ussl) },
modussl: SSL socket wrapper module based on axTLS.
2015-10-06 11:10:00 -04:00
#endif
compiles and runs; hangs on import storage;storage.VfsFat.<tab>
2018-07-13 22:51:10 -04:00
#if MICROPY_PY_LWIP
py: Add MP_ROM_* macros and mp_rom_* types and use them.
2015-11-27 08:38:15 -05:00
{ MP_ROM_QSTR(MP_QSTR_lwip), MP_ROM_PTR(&mp_module_lwip) },
extmod/modlwip: slip: Use stream protocol and be port-independent. Based on the original patch by Galen Hazelwood: https://github.com/micropython/micropython/pull/1517 .
2015-10-26 17:04:33 -04:00
#endif
extmod/modwebsocket: Start module for WebSocket helper functions. Currently, only write support is implemented (of limited buffer size).
2016-03-24 13:14:12 -04:00
#if MICROPY_PY_WEBSOCKET
{ MP_ROM_QSTR(MP_QSTR_websocket), MP_ROM_PTR(&mp_module_websocket) },
#endif
compiles and runs; hangs on import storage;storage.VfsFat.<tab>
2018-07-13 22:51:10 -04:00
#if MICROPY_PY_WEBREPL
extmod/modwebrepl: Module to handle WebREPL protocol. While just a websocket is enough for handling terminal part of WebREPL, handling file transfer operations requires demultiplexing and acting upon, which is encapsulated in _webrepl class provided by this module, which wraps a websocket object.
2016-04-28 17:52:52 -04:00
{ MP_ROM_QSTR(MP_QSTR__webrepl), MP_ROM_PTR(&mp_module_webrepl) },
#endif
extmod: Add initial framebuf module.
2016-04-08 06:08:37 -04:00
#if MICROPY_PY_FRAMEBUF
{ MP_ROM_QSTR(MP_QSTR_framebuf), MP_ROM_PTR(&mp_module_framebuf) },
#endif
py: Support to build berkeley db 1.85 and "btree" module.
2016-06-14 17:52:45 -04:00
#if MICROPY_PY_BTREE
{ MP_ROM_QSTR(MP_QSTR_btree), MP_ROM_PTR(&mp_module_btree) },
#endif
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
// extra builtin modules as defined by a port
MICROPY_PORT_BUILTIN_MODULES
Only include uheap in debug builds.
2016-11-29 13:46:47 -05:00
Measure and report maximum stack usage. (#175) Add max stack usage tracking, visible via debug module ustack. Add separate cpp flag for enabling modules: MICROPY_DEBUG_MODULES
2017-07-23 15:32:05 -04:00
#if defined(MICROPY_DEBUG_MODULES) && defined(MICROPY_PORT_BUILTIN_DEBUG_MODULES)
Only include uheap in debug builds.
2016-11-29 13:46:47 -05:00
, MICROPY_PORT_BUILTIN_DEBUG_MODULES
#endif
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
};
atmel-samd: Add help() and remove mem_info functions from smallest builds to make room.
2016-12-19 19:38:35 -05:00
MP_DEFINE_CONST_MAP(mp_builtin_module_map, mp_builtin_module_table);
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
py: Move weak-link map to objmodule.c, and expose module maps as public.
2017-01-19 18:21:30 -05:00
#if MICROPY_MODULE_WEAK_LINKS
STATIC const mp_rom_map_elem_t mp_builtin_module_weak_links_table[] = {
MICROPY_PORT_BUILTIN_MODULE_WEAK_LINKS
};
Proper support for registering builtin modules in ROM. Comes with some refactoring of code and renaming of files. All modules are now named mod*.[ch].
2014-03-25 10:18:18 -04:00
py: Move weak-link map to objmodule.c, and expose module maps as public.
2017-01-19 18:21:30 -05:00
MP_DEFINE_CONST_MAP(mp_builtin_module_weak_links_map, mp_builtin_module_weak_links_table);
#endif
Proper support for registering builtin modules in ROM. Comes with some refactoring of code and renaming of files. All modules are now named mod*.[ch].
2014-03-25 10:18:18 -04:00
// returns MP_OBJ_NULL if not found
mp_obj_t mp_module_get(qstr module_name) {
py/modsys: Implement sys.modules. This for example will allow people to reload modules which didn't load successfully (e.g. due to syntax error).
2015-12-04 17:09:10 -05:00
mp_map_t *mp_loaded_modules_map = &MP_STATE_VM(mp_loaded_modules_dict).map;
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
// lookup module
py/modsys: Implement sys.modules. This for example will allow people to reload modules which didn't load successfully (e.g. due to syntax error).
2015-12-04 17:09:10 -05:00
mp_map_elem_t *el = mp_map_lookup(mp_loaded_modules_map, MP_OBJ_NEW_QSTR(module_name), MP_MAP_LOOKUP);
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
py: Make globals and locals proper dictionary objects. Finishes addressing issue #424. In the end this was a very neat refactor that now makes things a lot more consistent across the py code base. It allowed some simplifications in certain places, now that everything is a dict object. Also converted builtins tables to dictionaries. This will be useful when we need to turn builtins into a proper module.
2014-04-05 17:36:42 -04:00
if (el == NULL) {
// module not found, look for builtin module names
py: Allow builtins to be overridden. This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS) which, when enabled, allows to override all names within the builtins module. A builtins override dict is created the first time the user assigns to a name in the builtins model, and then that dict is searched first on subsequent lookups. Note that this implementation doesn't allow deleting of names. This patch also does some refactoring of builtins code, creating the modbuiltins.c file. Addresses issue #959.
2014-12-09 11:19:48 -05:00
el = mp_map_lookup((mp_map_t*)&mp_builtin_module_map, MP_OBJ_NEW_QSTR(module_name), MP_MAP_LOOKUP);
py: Make globals and locals proper dictionary objects. Finishes addressing issue #424. In the end this was a very neat refactor that now makes things a lot more consistent across the py code base. It allowed some simplifications in certain places, now that everything is a dict object. Also converted builtins tables to dictionaries. This will be useful when we need to turn builtins into a proper module.
2014-04-05 17:36:42 -04:00
if (el == NULL) {
return MP_OBJ_NULL;
}
py/objmodule: Factor common code for calling __init__ on builtin module.
2018-02-20 01:56:58 -05:00
mp_module_call_init(module_name, el->value);
Implement modules as singletons Python semantics. In Python, importing module several times returns same underlying module object. This also fixes import statement handling for builtin modules. There're still issues: 1. CPython exposes set of loaded modules as sys.modules, we may want to do that either. 2. Builtin modules are implicitly imported, which is not really correct. We should separate registering a (builtin) module and importing a module. CPython keeps builtin module names in sys.builtin_module_names .
2014-01-19 17:03:34 -05:00
}
Implement ROMable modules. Add math module. mp_module_obj_t can now be put in ROM. Configuration of float type is now similar to longint: can now choose none, float or double as the implementation. math module has basic math functions. For STM port, these are not yet implemented (they are just stub functions).
2014-03-08 10:24:39 -05:00
py: Make globals and locals proper dictionary objects. Finishes addressing issue #424. In the end this was a very neat refactor that now makes things a lot more consistent across the py code base. It allowed some simplifications in certain places, now that everything is a dict object. Also converted builtins tables to dictionaries. This will be useful when we need to turn builtins into a proper module.
2014-04-05 17:36:42 -04:00
// module found, return it
return el->value;
Add mp_obj_module_register * Add function to load static modules. * Use module_register to pyb module.
2014-03-24 18:25:27 -04:00
}
py, unix, stmhal: Allow to compile with -Wshadow. See issue #699.
2015-01-20 06:55:10 -05:00
void mp_module_register(qstr qst, mp_obj_t module) {
py/modsys: Implement sys.modules. This for example will allow people to reload modules which didn't load successfully (e.g. due to syntax error).
2015-12-04 17:09:10 -05:00
mp_map_t *mp_loaded_modules_map = &MP_STATE_VM(mp_loaded_modules_dict).map;
mp_map_lookup(mp_loaded_modules_map, MP_OBJ_NEW_QSTR(qst), MP_MAP_LOOKUP_ADD_IF_NOT_FOUND)->value = module;
Add module object, to be used eventually for import.
2014-01-02 16:30:26 -05:00
}
py/objmodule: Factor common code for calling __init__ on builtin module.
2018-02-20 01:56:58 -05:00
#if MICROPY_MODULE_BUILTIN_INIT
void mp_module_call_init(qstr module_name, mp_obj_t module_obj) {
// Look for __init__ and call it if it exists
mp_obj_t dest[2];
mp_load_method_maybe(module_obj, MP_QSTR___init__, dest);
if (dest[0] != MP_OBJ_NULL) {
mp_call_method_n_kw(0, 0, dest);
// Register module so __init__ is not called again.
// If a module can be referenced by more than one name (eg due to weak links)
// then __init__ will still be called for each distinct import, and it's then
// up to the particular module to make sure it's __init__ code only runs once.
mp_module_register(module_name, module_obj);
}
}
#endif
Reference in New Issue Copy Permalink