everywhere: replace double words

import os
import re

common_words = set([
    'about', 'after', 'all', 'also', 'an', 'and',
     'any', 'are', 'as', 'at',
    'be', 'because', 'but', 'by', 'can', 'come',
    'could', 'day', 'do', 'even',
    'first', 'for', 'get', 'give', 'go', 'has',
    'have', 'he', 'her',
    'him', 'his', 'how', 'I', 'in', 'into', 'it',
    'its', 'just',
    'know', 'like', 'look', 'make', 'man', 'many',
    'me', 'more', 'my', 'new',
    'no', 'not', 'now', 'of', 'one', 'only', 'or',
    'other', 'our', 'out',
    'over', 'people', 'say', 'see', 'she', 'so',
    'some', 'take', 'tell', 'than',
    'their', 'them', 'then', 'there', 'these',
    'they', 'think',
    'this', 'time', 'two', 'up', 'use', 'very',
    'want', 'was', 'way',
    'we', 'well', 'what', 'when', 'which', 'who',
    'will', 'with', 'would',
    'year', 'you', 'your'
])

valid_extensions = set([
    'c', 'h', 'yaml', 'cmake', 'conf', 'txt', 'overlay',
    'rst', 'dtsi',
    'Kconfig', 'dts', 'defconfig', 'yml', 'ld', 'sh', 'py',
    'soc', 'cfg'
])

def filter_repeated_words(text):
    # Split the text into lines
    lines = text.split('\n')

    # Combine lines into a single string with unique separator
    combined_text = '/*sep*/'.join(lines)

    # Replace repeated words within a line
    def replace_within_line(match):
        return match.group(1)

    # Regex for matching repeated words within a line
    within_line_pattern =
	re.compile(r'\b(' +
		'|'.join(map(re.escape, common_words)) +
		r')\b\s+\b\1\b')
    combined_text = within_line_pattern.
		sub(replace_within_line, combined_text)

    # Replace repeated words across line boundaries
    def replace_across_lines(match):
        return match.group(1) + match.group(2)

    # Regex for matching repeated words across line boundaries
    across_lines_pattern = re.
		compile(r'\b(' + '|'.join(
			map(re.escape, common_words)) +
			r')\b(\s*[*\/\n\s]*)\b\1\b')
    combined_text = across_lines_pattern.
		sub(replace_across_lines, combined_text)

    # Split the text back into lines
    filtered_text = combined_text.split('/*sep*/')

    return '\n'.join(filtered_text)

def process_file(file_path):
    with open(file_path, 'r', encoding='utf-8') as file:
        text = file.read()

    new_text = filter_repeated_words(text)

    with open(file_path, 'w', encoding='utf-8') as file:
        file.write(new_text)

def process_directory(directory_path):
    for root, dirs, files in os.walk(directory_path):
        dirs[:] = [d for d in dirs if not d.startswith('.')]
        for file in files:
            # Filter out hidden files
            if file.startswith('.'):
                continue
            file_extension = file.split('.')[-1]
            if
	file_extension in valid_extensions:  # 只处理指定后缀的文件
                file_path = os.path.join(root, file)
                print(f"Processed file: {file_path}")
                process_file(file_path)

directory_to_process = "/home/mi/works/github/zephyrproject/zephyr"
process_directory(directory_to_process)

Signed-off-by: Lingao Meng <menglingao@xiaomi.com>

arch/arc/core/mpu/arc_mpu_v2_internal.h

@@ -118,7 +118,7 @@ static inline bool _is_enabled_region(uint32_t r_index)
 }
 
 /**
- * This internal function check if the given buffer in in the region
+ * This internal function check if the given buffer in the region
  */
 static inline bool _is_in_region(uint32_t r_index, uint32_t start, uint32_t size)
 {

arch/arc/core/mpu/arc_mpu_v6_internal.h

@@ -156,7 +156,7 @@ static inline bool _is_enabled_region(uint32_t r_index)
 }
 
 /**
- * This internal function check if the given buffer in in the region
+ * This internal function check if the given buffer in the region
  */
 static inline bool _is_in_region(uint32_t r_index, uint32_t start, uint32_t size)
 {

arch/xtensa/core/README_MMU.txt

@@ -254,7 +254,7 @@ of access variability. But it also means that the TLB entries end up
 being stored twice in the same CPU, wasting transistors that could
 presumably store other useful data.
 
-But it it also important to note that the L1 data cache on Xtensa is
+But it also important to note that the L1 data cache on Xtensa is
 incoherent! The cache being used for refill reflects the last access
 on the current CPU only, and not of the underlying memory being
 mapped. Page table changes in the data cache of one CPU will be

boards/native/native_posix/native_rtc.h

@@ -44,7 +44,7 @@ extern "C" {
 uint64_t native_rtc_gettime_us(int clock_type);
 
 /**
- * @brief Get the value of a clock split in in nsec and seconds
+ * @brief Get the value of a clock split in nsec and seconds
  *
  * @param clock_type Which clock to measure from
  * @param nsec Pointer to store the nanoseconds

doc/connectivity/networking/api/http_server.rst

@@ -103,7 +103,7 @@ macro:
 
  HTTP_SERVICE_DEFINE(my_service, "0.0.0.0", &http_service_port, 1, 10, NULL);
 
-Alternatively, an HTTPS service can be defined with with
+Alternatively, an HTTPS service can be defined with
 :c:macro:`HTTPS_SERVICE_DEFINE`:
 
 .. code-block:: c

doc/connectivity/networking/api/lwm2m.rst

@@ -670,7 +670,7 @@ where it cannot recover.
  - Try to recover network connection. Then restart the client by calling :c:func:`lwm2m_rd_client_start`.
  This might also indicate configuration issue.
 
-Sending of data in the table above refers to calling :c:func:`lwm2m_send_cb` or by writing into of of the observed resources where observation would trigger a notify message.
+Sending of data in the table above refers to calling :c:func:`lwm2m_send_cb` or by writing into of the observed resources where observation would trigger a notify message.
 Receiving of data refers to receiving read, write or execute operations from the server. Application can register callbacks for these operations.
 
 Configuring lifetime and activity period

doc/connectivity/networking/net_config_guide.rst

@@ -174,7 +174,7 @@ TCP Options
 :kconfig:option:`CONFIG_NET_TCP_RETRY_COUNT`
  Maximum number of TCP segment retransmissions.
  The following formula can be used to determine the time (in ms)
- that a segment will be be buffered awaiting retransmission:
+ that a segment will be buffered awaiting retransmission:
 
  .. math::
 

doc/develop/flash_debug/probes.rst

@@ -210,7 +210,7 @@ The probe may be updated to use CMSIS-DAP firmware with the following steps:
  be done by installing :ref:`linkserver-debug-host-tools`.
 
 #. Put the LPC-Link2 microcontroller into DFU boot mode by attaching the DFU
- jumper, then then connecting to the USB debug port on the board. This
+ jumper, then connecting to the USB debug port on the board. This
  jumper is connected to ``P2_6`` on the LPC4322 SOC.
 
 #. Run the ``program_CMSIS`` script, found in the installed LPCScrypt ``scripts``
@@ -241,7 +241,7 @@ The probe may be updated to use the J-Link firmware with the following steps:
  be done by installing :ref:`linkserver-debug-host-tools`.
 
 #. Put the LPC-Link2 microcontroller into DFU boot mode by attaching the DFU
- jumper, then then connecting to the USB debug port on the board. This
+ jumper, then connecting to the USB debug port on the board. This
  jumper is connected to ``P2_6`` on the LPC4322 SOC.
 
 #. Run the ``program_JLINK`` script, found in the installed LPCScrypt ``scripts``

doc/releases/release-notes-1.14.rst

@@ -2067,7 +2067,7 @@ release:
 * :github:`10345` - The OpenAMP remote build is for wrong board
 * :github:`10344` - SPI Chip Select usage is not unambiguous
 * :github:`10329` - SystemView overflow event
-* :github:`10320` - arm: mpu: mpu_config and and mpu_regions to be declared/defined as const
+* :github:`10320` - arm: mpu: mpu_config and mpu_regions to be declared/defined as const
 * :github:`10318` - It is not documented what YAML bindings do
 * :github:`10316` - net: sockets: Close doesn't unblock recv
 * :github:`10313` - net: sockets: Packets are leaked on TCP abort connection

doc/releases/release-notes-3.1.rst

@@ -490,7 +490,7 @@ Drivers and Sensors
  series. Interrupt driven mode. Supports 1 and 8 lines in Single or Dual
  Transfer Modes.
  * STM32L5: Added support for Single Bank.
- * STM32 QSPI driver was extended with with QER (SFDP, DTS), custom quad write opcode
+ * STM32 QSPI driver was extended with QER (SFDP, DTS), custom quad write opcode
  and 1-1-4 read mode.
  * Added support for STM32U5 series.
 

doc/releases/release-notes-3.3.rst

@@ -210,7 +210,7 @@ Deprecated in this release
 
 * MCUmgr subsystem, specifically the SMP transport API, is dropping `zephyr_`
  prefix, deprecating prefixed functions and callback type definitions with the
- prefix and replacing them with with prefix-less variants.
+ prefix and replacing them with prefix-less variants.
  The :c:struct:`zephyr_smp_transport` type, representing transport object,
  is now replaced with :c:struct:`smp_transport`, and the later one is used,
  instead of the former one, by all prefix-less functions.
@@ -3486,7 +3486,7 @@ Addressed issues
 * :github:`50732` - net: tests/net/ieee802154/l2/net.ieee802154.l2 failed on reel_board due to build failure
 * :github:`50709` - tests: arch: arm: arm_thread_swap fails on stm32g0 or stm32l0
 * :github:`50684` - After enabling CONFIG_SPI_STM32_DMA in project config file for STM32MP157-dk2 Zephyr throwing error
-* :github:`50665` - MEC15xx/MEC1501: UART and and special purpose pins missing pinctrl configuration
+* :github:`50665` - MEC15xx/MEC1501: UART and special purpose pins missing pinctrl configuration
 * :github:`50658` - Bluetooth: BLE stack notifications blocks host side for too long (``drivers/bluetooth/hci/spi.c`` and ``hci_spi``)
 * :github:`50656` - Wrong definition of bank size for intel memory management driver.
 * :github:`50655` - STM32WB55 Bus Fault when connecting then disconnecting then connecting then disconnecting then connecting

doc/services/mem_mgmt/index.rst

@@ -92,7 +92,7 @@ to define and create a set of memory heaps from which the user can allocate
 memory from with certain attributes / capabilities.
 
 When the :kconfig:option:`CONFIG_MEM_ATTR_HEAP` is set, every region marked
-with one of the memory attributes listed in in
+with one of the memory attributes listed in
 :zephyr_file:`include/zephyr/dt-bindings/memory-attr/memory-attr-sw.h` is added
 to a pool of memory heaps used for dynamic allocation of memory buffers with
 certain attributes.

drivers/adc/adc_ads1x1x.c

@@ -616,7 +616,7 @@ static const struct adc_driver_api ads1x1x_api = {
 /* The ADS111X provides 16 bits of data in binary two's complement format
  * A positive full-scale (+FS) input produces an output code of 7FFFh and a
  * negative full-scale (–FS) input produces an output code of 8000h. Single
- * ended signal measurements only only use the positive code range from
+ * ended signal measurements only use the positive code range from
  * 0000h to 7FFFh
  */
 #define ADS111X_RESOLUTION 16
@@ -659,7 +659,7 @@ DT_INST_FOREACH_STATUS_OKAY(ADS1113_INIT)
 /* The ADS101X provides 12 bits of data in binary two's complement format
  * A positive full-scale (+FS) input produces an output code of 7FFh and a
  * negative full-scale (–FS) input produces an output code of 800h. Single
- * ended signal measurements only only use the positive code range from
+ * ended signal measurements only use the positive code range from
  * 000h to 7FFh
  */
 #define ADS101X_RESOLUTION 12

drivers/audio/dmic_nrfx_pdm.c

@@ -207,7 +207,7 @@ static bool check_pdm_frequencies(const struct dmic_nrfx_pdm_drv_cfg *drv_cfg,
  better_found = true;
  }
 
- /* Since frequencies are are in ascending order, stop
+ /* Since frequencies are in ascending order, stop
  * checking next ones for the current ratio after
  * resulting PCM rate goes above the one requested.
  */

drivers/bluetooth/hci/Kconfig

@@ -160,7 +160,7 @@ config BT_SPI_INIT_PRIORITY
  default 75
 
 config BT_BLUENRG_ACI
- bool "ACI message with with BlueNRG-based devices"
+ bool "ACI message with BlueNRG-based devices"
  select BT_HCI_SET_PUBLIC_ADDR
  help
  Enable support for devices compatible with the BlueNRG Bluetooth

drivers/clock_control/clock_control_mchp_xec.c

@@ -267,7 +267,7 @@ static int periph_clk_src_using_pin(enum periph_clk32k_src src)
 /* MEC15xx uses the same 32KHz source for both PLL and Peripheral 32K clock domains.
  * We ignore the peripheral clock source.
  * If XTAL is selected (parallel) or single-ended the external 32KHz MUST stay on
- * even when when VTR goes off.
+ * even when VTR goes off.
  * If PIN(32KHZ_IN pin) as the external source, hardware can auto-switch to internal
  * silicon OSC if the signal on the 32KHZ_PIN goes away.
  * We ignore th

drivers/dma/dma_xmc4xxx.c

@@ -393,7 +393,7 @@ static int dma_xmc4xxx_get_status(const struct device *dev, uint32_t channel,
 
  stat->pending_length = dma_channel->block_ts - XMC_DMA_CH_GetTransferredData(dma, channel);
  stat->pending_length *= dma_channel->source_data_size;
- /* stat->dir and other remaining fields are not set. They are are not */
+ /* stat->dir and other remaining fields are not set. They are not */
  /* useful for xmc4xxx peripheral drivers. */
 
  return 0;

drivers/ethernet/eth_smsc91x.c

@@ -522,7 +522,7 @@ static int smsc_send_pkt(struct smsc_data *sc, uint8_t *buf, uint16_t len)
  smsc_write_2(sc, DATA0, len + PKT_CTRL_DATA_LEN);
  smsc_write_multi_2(sc, DATA0, (uint16_t *)buf, len / 2);
 
- /* Push out the control byte and and the odd byte if needed. */
+ /* Push out the control byte and the odd byte if needed. */
  if (len & 1) {
  smsc_write_2(sc, DATA0, (CTRL_ODD << 8) | buf[len - 1]);
  } else {

drivers/sdhc/rcar_mmc.c

@@ -2045,7 +2045,7 @@ static int rcar_mmc_init_controller_regs(const struct device *dev)
  reg |= RCAR_MMC_DMA_MODE_ADDR_INC | RCAR_MMC_DMA_MODE_WIDTH;
  rcar_mmc_write_reg32(dev, RCAR_MMC_DMA_MODE, reg);
 
- /* store version of of introductory IP */
+ /* store version of introductory IP */
  data->ver = rcar_mmc_read_reg32(dev, RCAR_MMC_VERSION);
  data->ver &= RCAR_MMC_VERSION_IP;
 

drivers/sensor/bosch/bme280/bme280.c

@@ -121,7 +121,7 @@ static int bme280_wait_until_ready(const struct device *dev)
  uint8_t status = 0;
  int ret;
 
- /* Wait for NVM to copy and and measurement to be completed */
+ /* Wait for NVM to copy and measurement to be completed */
  do {
  k_sleep(K_MSEC(3));
  ret = bme280_reg_read(dev, BME280_REG_STATUS, &status, 1);

drivers/serial/leuart_gecko.c

@@ -65,7 +65,7 @@ static void leuart_gecko_poll_out(const struct device *dev, unsigned char c)
  LEUART_TypeDef *base = DEV_BASE(dev);
 
  /* LEUART_Tx function already waits for the transmit buffer being empty
- * and and waits for the bus to be free to transmit.
+ * and waits for the bus to be free to transmit.
  */
  LEUART_Tx(base, c);
 }

dts/bindings/pinctrl/ti,cc13xx-cc26xx-pinctrl.yaml

@@ -123,7 +123,7 @@ child-binding:
  configured as an output.
  2: min 2 mA (SoC default)
  4: min 4 mA
- 8: min 8 mA for for double drive strength IOs, min 4 mA for normal IOs
+ 8: min 8 mA for double drive strength IOs, min 4 mA for normal IOs
 
  ti,input-edge-detect:
  type: int

include/zephyr/bluetooth/audio/media_proxy.h

@@ -787,7 +787,7 @@ int media_proxy_ctrl_get_track_position(struct media_player *player);
  * @brief Set Track Position
  *
  * Set the playing position of the media player in the current
- * track. The position is given in in hundredths of a second,
+ * track. The position is given in hundredths of a second,
  * from the beginning of the track of the track for positive
  * values, and (backwards) from the end of the track for
  * negative values.
@@ -1178,7 +1178,7 @@ struct media_proxy_pl_calls {
  * @brief Set Track Position
  *
  * Set the playing position of the media player in the current
- * track. The position is given in in hundredths of a second,
+ * track. The position is given in hundredths of a second,
  * from the beginning of the track of the track for positive
  * values, and (backwards) from the end of the track for
  * negative values.

include/zephyr/drivers/i2c/rtio.h

@@ -43,7 +43,7 @@ struct i2c_rtio {
  };
 
 /**
- * @brief Copy an an array of i2c_msgs to rtio submissions and a transaction
+ * @brief Copy an array of i2c_msgs to rtio submissions and a transaction
  *
  * @retval sqe Last sqe setup in the copy
  * @retval NULL Not enough memory to copy the transaction

include/zephyr/kernel/mm/demand_paging.h

@@ -252,7 +252,7 @@ void k_mem_paging_eviction_remove(struct k_mem_page_frame *pf);
  *
  * The architecture-specific memory fault handler will invoke this to tell the
  * eviction algorithm the provided physical address belongs to a page frame
- * being accessed and such page frame should become unlikely to be be
+ * being accessed and such page frame should become unlikely to be
  * considered as the next eviction candidate.
  *
  * This function is invoked with interrupts locked.

include/zephyr/net/net_if.h

@@ -2985,7 +2985,7 @@ static inline void net_if_unset_promisc(struct net_if *iface)
  * @param iface Pointer to network interface
  *
  * @return True if interface is in promisc mode,
- * False if interface is not in in promiscuous mode.
+ * False if interface is not in promiscuous mode.
  */
 #if defined(CONFIG_NET_PROMISCUOUS_MODE)
 bool net_if_is_promisc(struct net_if *iface);

include/zephyr/sys/onoff.h

@@ -340,7 +340,7 @@ static inline bool onoff_has_error(const struct onoff_manager *mgr)
  *
  * @retval non-negative the observed state of the machine at the time
  * the request was processed, if successful.
- * @retval -EIO if service has recorded an an error.
+ * @retval -EIO if service has recorded an error.
  * @retval -EINVAL if the parameters are invalid.
  * @retval -EAGAIN if the reference count would overflow.
  */
@@ -361,7 +361,7 @@ int onoff_request(struct onoff_manager *mgr,
  *
  * @retval non-negative the observed state (ONOFF_STATE_ON) of the
  * machine at the time of the release, if the release succeeds.
- * @retval -EIO if service has recorded an an error.
+ * @retval -EIO if service has recorded an error.
  * @retval -ENOTSUP if the machine is not in a state that permits
  * release.
  */

include/zephyr/sys/sys_heap.h

@@ -42,7 +42,7 @@ extern "C" {
  * are constant time (though there is a search of the smallest bucket
  * that has a compile-time-configurable upper bound, setting this to
  * extreme values results in an effectively linear search of the
- * list), objectively fast (~hundred instructions) and and amenable to
+ * list), objectively fast (~hundred instructions) and amenable to
  * locked operation.
  */
 

kernel/Kconfig

@@ -299,7 +299,7 @@ choice SCHED_ALGORITHM
  prompt "Scheduler priority queue algorithm"
  default SCHED_DUMB
  help
- The kernel can be built with with several choices for the
+ The kernel can be built with several choices for the
  ready queue implementation, offering different choices between
  code size, constant factor runtime overhead and performance
  scaling when many threads are added.

kernel/userspace.c

@@ -82,7 +82,7 @@ static void clear_perms_cb(struct k_object *ko, void *ctx_ptr);
 const char *otype_to_str(enum k_objects otype)
 {
  const char *ret;
- /* -fdata-sections doesn't work right except in very very recent
+ /* -fdata-sections doesn't work right except in very recent
  * GCC and these literal strings would appear in the binary even if
  * otype_to_str was omitted by the linker
  */

samples/subsys/usb/webusb/src/main.c

@@ -64,7 +64,7 @@ static struct msosv2_descriptor_t {
  .wLength = sizeof(struct msosv2_function_subset_header),
  .wDescriptorType = MS_OS_20_SUBSET_HEADER_FUNCTION,
  /* The WebUSB interface number becomes the first when CDC_ACM is enabled by
- * configuration. Beware that if this sample is used as as inspiration for
+ * configuration. Beware that if this sample is used as inspiration for
  * applications, where the WebUSB interface is no longer the first,
  * remember to adjust bFirstInterface.
  */

samples/tfm_integration/psa_crypto/src/psa_attestation.c

@@ -33,7 +33,7 @@ psa_status_t att_get_iat(uint8_t *ch_buffer, uint32_t ch_sz,
  size_t token_buf_size = ATT_MAX_TOKEN_SIZE;
 
 
- /* Call with with bigger challenge object than allowed */
+ /* Call with bigger challenge object than allowed */
 
  /*
  * First determine how large the token is on this system.

scripts/build/subfolder_list.py

@@ -30,7 +30,7 @@ def parse_args():
  help='File to write containing a list of all \
  directories found')
  parser.add_argument('-t', '--trigger-file', required=False,
- help='Trigger file to be be touched to re-run CMake')
+ help='Trigger file to be touched to re-run CMake')
 
  args = parser.parse_args()
 

scripts/kconfig/kconfiglib.py

@@ -3729,7 +3729,7 @@ def _add_props_to_sym(self, node):
 
  if node.is_configdefault:
  # Store any defaults for later application after the complete tree
- # is known. The current length of of the default array is stored so
+ # is known. The current length of the default array is stored so
  # the configdefaults can be inserted in the order they originally
  # appeared.
  sym.configdefaults.append((len(sym.defaults), node.defaults))

scripts/native_simulator/native/src/include/native_rtc.h

@@ -43,7 +43,7 @@ extern "C" {
 uint64_t native_rtc_gettime_us(int clock_type);
 
 /**
- * @brief Get the value of a clock split in in nsec and seconds
+ * @brief Get the value of a clock split in nsec and seconds
  *
  * @param clock_type Which clock to measure from
  * @param nsec Pointer to store the nanoseconds

soc/ite/ec/common/chip_chipregs.h

@@ -615,7 +615,7 @@ struct ep_ext_regs_9x {
 };
 
 /* From BXh to BDh are EP FIFO 1-3 Control 0/1 Registers, and their
- * definitions as as follows:
+ * definitions as follows:
  * B8h: EP_FIFO1_CONTROL0_REG
  * B9h: EP_FIFO1_CONTROL1_REG
  * BAh: EP_FIFO2_CONTROL0_REG
@@ -657,7 +657,7 @@ struct ep_ext_regs_bx {
 
 
 /* From D6h to DDh are EP Extended Control Registers, and their
- * definitions as as follows:
+ * definitions as follows:
  * D6h: EP0_EXT_CTRL1
  * D7h: EP0_EXT_CTRL2
  * D8h: EP1_EXT_CTRL1

subsys/bluetooth/audio/shell/media_controller.c

@@ -603,7 +603,7 @@ static int cmd_media_read_playback_speed(const struct shell *sh, size_t argc, ch
  int err = media_proxy_ctrl_get_playback_speed(current_player);
 
  if (err) {
- shell_error(ctx_shell, "Playback speed get get failed (%d)", err);
+ shell_error(ctx_shell, "Playback speed get failed (%d)", err);
  }
 
  return err;