neovim

buffer.c (46605B)

---

// Some of this code was adapted from 'if_py_both.h' from the original
// vim source

#include <lauxlib.h>
#include <stdbool.h>
#include <stddef.h>
#include <stdint.h>
#include <string.h>

#include "klib/kvec.h"
#include "lua.h"
#include "nvim/api/buffer.h"
#include "nvim/api/keysets_defs.h"
#include "nvim/api/private/defs.h"
#include "nvim/api/private/dispatch.h"
#include "nvim/api/private/helpers.h"
#include "nvim/api/private/validate.h"
#include "nvim/ascii_defs.h"
#include "nvim/autocmd.h"
#include "nvim/autocmd_defs.h"
#include "nvim/buffer.h"
#include "nvim/buffer_defs.h"
#include "nvim/buffer_updates.h"
#include "nvim/change.h"
#include "nvim/cursor.h"
#include "nvim/ex_cmds.h"
#include "nvim/extmark.h"
#include "nvim/extmark_defs.h"
#include "nvim/globals.h"
#include "nvim/lua/executor.h"
#include "nvim/mapping.h"
#include "nvim/mark.h"
#include "nvim/mark_defs.h"
#include "nvim/marktree_defs.h"
#include "nvim/memline.h"
#include "nvim/memline_defs.h"
#include "nvim/memory.h"
#include "nvim/memory_defs.h"
#include "nvim/move.h"
#include "nvim/ops.h"
#include "nvim/option_vars.h"
#include "nvim/pos_defs.h"
#include "nvim/state_defs.h"
#include "nvim/types_defs.h"
#include "nvim/undo.h"
#include "nvim/undo_defs.h"
#include "nvim/vim_defs.h"

#include "api/buffer.c.generated.h"

/// Ensures that a buffer is loaded.
buf_T *api_buf_ensure_loaded(Buffer buffer, Error *err)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);
 if (!buf) {
   return NULL;
 }

 if (buf->b_ml.ml_mfp == NULL && !buf_ensure_loaded(buf)) {
   api_set_error(err, kErrorTypeException, "Failed to load buffer");
   return NULL;
 }

 return buf;
}

/// @brief <pre>help
/// For more information on buffers, see |buffers|.
///
/// Unloaded Buffers: ~
///
/// Buffers may be unloaded by the |:bunload| command or the buffer's
/// 'bufhidden' option. When a buffer is unloaded its file contents are freed
/// from memory and vim cannot operate on the buffer lines until it is reloaded
/// (usually by opening the buffer again in a new window). API methods such as
/// |nvim_buf_get_lines()| and |nvim_buf_line_count()| will be affected.
///
/// You can use |nvim_buf_is_loaded()| or |nvim_buf_line_count()| to check
/// whether a buffer is loaded.
/// </pre>

/// Returns the number of lines in the given buffer.
///
/// @param buffer   Buffer id, or 0 for current buffer
/// @param[out] err Error details, if any
/// @return Line count, or 0 for unloaded buffer. |api-buffer|
Integer nvim_buf_line_count(Buffer buffer, Error *err)
 FUNC_API_SINCE(1)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return 0;
 }

 // return sentinel value if the buffer isn't loaded
 if (buf->b_ml.ml_mfp == NULL) {
   return 0;
 }

 return buf->b_ml.ml_line_count;
}

/// Activates |api-buffer-updates| events on a channel, or as Lua callbacks.
///
/// Example (Lua): capture buffer updates in a global `events` variable
/// (use "vim.print(events)" to see its contents):
///
/// ```lua
/// events = {}
/// vim.api.nvim_buf_attach(0, false, {
///   on_lines = function(...)
///     table.insert(events, {...})
///   end,
/// })
/// ```
///
/// @see |nvim_buf_detach()|
/// @see |api-buffer-updates-lua|
///
/// @param channel_id
/// @param buffer Buffer id, or 0 for current buffer
/// @param send_buffer True if the initial notification should contain the
///        whole buffer: first notification will be `nvim_buf_lines_event`.
///        Else the first notification will be `nvim_buf_changedtick_event`.
///        Not for Lua callbacks.
/// @param  opts  Optional parameters.
///             - on_lines: Called on linewise changes. Not called on buffer reload (`:checktime`,
///               `:edit`, …), see `on_reload:`. Return a [lua-truthy] value to detach. Args:
///               - the string "lines"
///               - buffer id
///               - b:changedtick
///               - first line that changed (zero-indexed)
///               - last line that was changed
///               - last line in the updated range
///               - byte count of previous contents
///               - deleted_codepoints (if `utf_sizes` is true)
///               - deleted_codeunits (if `utf_sizes` is true)
///             - on_bytes: Called on granular changes (compared to on_lines). Not called on buffer
///               reload (`:checktime`, `:edit`, …), see `on_reload:`. Return a [lua-truthy] value
///               to detach. Args:
///               - the string "bytes"
///               - buffer id
///               - b:changedtick
///               - start row of the changed text (zero-indexed)
///               - start column of the changed text
///               - byte offset of the changed text (from the start of
///                   the buffer)
///               - old end row of the changed text (offset from start row)
///               - old end column of the changed text
///                 (if old end row = 0, offset from start column)
///               - old end byte length of the changed text
///               - new end row of the changed text (offset from start row)
///               - new end column of the changed text
///                 (if new end row = 0, offset from start column)
///               - new end byte length of the changed text
///             - on_changedtick: Called on [changetick] increment without text change. Args:
///               - the string "changedtick"
///               - buffer id
///               - b:changedtick
///             - on_detach: Called on detach. Args:
///               - the string "detach"
///               - buffer id
///             - on_reload: Called on whole-buffer load (`:checktime`, `:edit`, …). Clients should
///               typically re-fetch the entire buffer contents. Args:
///               - the string "reload"
///               - buffer id
///             - utf_sizes: include UTF-32 and UTF-16 size of the replaced
///               region, as args to `on_lines`.
///             - preview: also attach to command preview (i.e. 'inccommand')
///               events.
/// @param[out] err Error details, if any
/// @return False if attach failed (invalid parameter, or buffer isn't loaded);
///         otherwise True.
Boolean nvim_buf_attach(uint64_t channel_id, Buffer buffer, Boolean send_buffer,
                       Dict(buf_attach) *opts, Error *err)
 FUNC_API_SINCE(4)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return false;
 }

 BufUpdateCallbacks cb = BUF_UPDATE_CALLBACKS_INIT;

 if (channel_id == LUA_INTERNAL_CALL) {
   if (HAS_KEY(opts, buf_attach, on_lines)) {
     cb.on_lines = opts->on_lines;
     opts->on_lines = LUA_NOREF;
   }

   if (HAS_KEY(opts, buf_attach, on_bytes)) {
     cb.on_bytes = opts->on_bytes;
     opts->on_bytes = LUA_NOREF;
   }

   if (HAS_KEY(opts, buf_attach, on_changedtick)) {
     cb.on_changedtick = opts->on_changedtick;
     opts->on_changedtick = LUA_NOREF;
   }

   if (HAS_KEY(opts, buf_attach, on_detach)) {
     cb.on_detach = opts->on_detach;
     opts->on_detach = LUA_NOREF;
   }

   if (HAS_KEY(opts, buf_attach, on_reload)) {
     cb.on_reload = opts->on_reload;
     opts->on_reload = LUA_NOREF;
   }

   cb.utf_sizes = opts->utf_sizes;

   cb.preview = opts->preview;
 }

 return buf_updates_register(buf, channel_id, cb, send_buffer);
}

/// Deactivates buffer-update events on the channel.
///
/// @see |nvim_buf_attach()|
/// @see |api-lua-detach| for detaching Lua callbacks
///
/// @param channel_id
/// @param buffer Buffer id, or 0 for current buffer
/// @param[out] err Error details, if any
/// @return False if detach failed (because the buffer isn't loaded);
///         otherwise True.
Boolean nvim_buf_detach(uint64_t channel_id, Buffer buffer, Error *err)
 FUNC_API_SINCE(4) FUNC_API_REMOTE_ONLY
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return false;
 }

 buf_updates_unregister(buf, channel_id);
 return true;
}

/// Gets a line-range from the buffer.
///
/// Indexing is zero-based, end-exclusive. Negative indices are interpreted
/// as length+1+index: -1 refers to the index past the end. So to get the
/// last element use start=-2 and end=-1.
///
/// Out-of-bounds indices are clamped to the nearest valid value, unless
/// `strict_indexing` is set.
///
/// @see |nvim_buf_get_text()|
///
/// @param channel_id
/// @param buffer           Buffer id, or 0 for current buffer
/// @param start            First line index
/// @param end              Last line index, exclusive
/// @param strict_indexing  Whether out-of-bounds should be an error.
/// @param[out] err         Error details, if any
/// @return Array of lines, or empty array for unloaded buffer.
ArrayOf(String) nvim_buf_get_lines(uint64_t channel_id,
                                  Buffer buffer,
                                  Integer start,
                                  Integer end,
                                  Boolean strict_indexing,
                                  Arena *arena,
                                  lua_State *lstate,
                                  Error *err)
 FUNC_API_SINCE(1)
{
 Array rv = ARRAY_DICT_INIT;
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return rv;
 }

 // return sentinel value if the buffer isn't loaded
 if (buf->b_ml.ml_mfp == NULL) {
   return rv;
 }

 bool oob = false;
 start = normalize_index(buf, start, true, &oob);
 end = normalize_index(buf, end, true, &oob);

 VALIDATE((!strict_indexing || !oob), "%s", "Index out of bounds", {
   return rv;
 });

 if (start >= end) {
   // Return 0-length array
   return rv;
 }

 size_t size = (size_t)(end - start);

 init_line_array(lstate, &rv, size, arena);

 buf_collect_lines(buf, size, (linenr_T)start, 0, (channel_id != VIML_INTERNAL_CALL), &rv,
                   lstate, arena);

 return rv;
}

/// Sets (replaces) a line-range in the buffer.
///
/// Indexing is zero-based, end-exclusive. Negative indices are interpreted
/// as length+1+index: -1 refers to the index past the end. So to change
/// or delete the last line use start=-2 and end=-1.
///
/// To insert lines at a given index, set `start` and `end` to the same index.
/// To delete a range of lines, set `replacement` to an empty array.
///
/// Out-of-bounds indices are clamped to the nearest valid value, unless
/// `strict_indexing` is set.
///
/// @see |nvim_buf_set_text()|
///
/// @param channel_id
/// @param buffer           Buffer id, or 0 for current buffer
/// @param start            First line index
/// @param end              Last line index, exclusive
/// @param strict_indexing  Whether out-of-bounds should be an error.
/// @param replacement      Array of lines to use as replacement
/// @param[out] err         Error details, if any
void nvim_buf_set_lines(uint64_t channel_id, Buffer buffer, Integer start, Integer end,
                       Boolean strict_indexing, ArrayOf(String) replacement, Arena *arena,
                       Error *err)
 FUNC_API_SINCE(1)
 FUNC_API_TEXTLOCK_ALLOW_CMDWIN
{
 buf_T *buf = api_buf_ensure_loaded(buffer, err);

 if (!buf) {
   return;
 }

 bool oob = false;
 start = normalize_index(buf, start, true, &oob);
 end = normalize_index(buf, end, true, &oob);

 VALIDATE((!strict_indexing || !oob), "%s", "Index out of bounds", {
   return;
 });
 VALIDATE((start <= end), "%s", "'start' is higher than 'end'", {
   return;
 });

 bool disallow_nl = (channel_id != VIML_INTERNAL_CALL);
 if (!check_string_array(replacement, "replacement string", disallow_nl, err)) {
   return;
 }

 size_t new_len = replacement.size;
 size_t old_len = (size_t)(end - start);
 ptrdiff_t extra = 0;  // lines added to text, can be negative
 char **lines = (new_len != 0) ? arena_alloc(arena, new_len * sizeof(char *), true) : NULL;

 for (size_t i = 0; i < new_len; i++) {
   const String l = replacement.items[i].data.string;

   // Fill lines[i] with l's contents. Convert NULs to newlines as required by
   // NL-used-for-NUL.
   lines[i] = arena_memdupz(arena, l.data, l.size);
   memchrsub(lines[i], NUL, NL, l.size);
 }

 TRY_WRAP(err, {
   if (!MODIFIABLE(buf)) {
     api_set_error(err, kErrorTypeException, "Buffer is not 'modifiable'");
     goto end;
   }

   if (u_save_buf(buf, (linenr_T)(start - 1), (linenr_T)end) == FAIL) {
     api_set_error(err, kErrorTypeException, "Failed to save undo information");
     goto end;
   }

   bcount_t deleted_bytes = get_region_bytecount(buf, (linenr_T)start, (linenr_T)end, 0, 0);

   // If the size of the range is reducing (ie, new_len < old_len) we
   // need to delete some old_len. We do this at the start, by
   // repeatedly deleting line "start".
   size_t to_delete = (new_len < old_len) ? old_len - new_len : 0;
   for (size_t i = 0; i < to_delete; i++) {
     if (ml_delete_buf(buf, (linenr_T)start, false) == FAIL) {
       api_set_error(err, kErrorTypeException, "Failed to delete line");
       goto end;
     }
   }

   if (to_delete > 0) {
     extra -= (ptrdiff_t)to_delete;
   }

   // For as long as possible, replace the existing old_len with the
   // new old_len. This is a more efficient operation, as it requires
   // less memory allocation and freeing.
   size_t to_replace = old_len < new_len ? old_len : new_len;
   bcount_t inserted_bytes = 0;
   for (size_t i = 0; i < to_replace; i++) {
     int64_t lnum = start + (int64_t)i;

     VALIDATE(lnum < MAXLNUM, "%s", "Index out of bounds", {
       goto end;
     });

     if (ml_replace_buf(buf, (linenr_T)lnum, lines[i], false, true) == FAIL) {
       api_set_error(err, kErrorTypeException, "Failed to replace line");
       goto end;
     }

     inserted_bytes += (bcount_t)strlen(lines[i]) + 1;
   }

   // Now we may need to insert the remaining new old_len
   for (size_t i = to_replace; i < new_len; i++) {
     int64_t lnum = start + (int64_t)i - 1;

     VALIDATE(lnum < MAXLNUM, "%s", "Index out of bounds", {
       goto end;
     });

     if (ml_append_buf(buf, (linenr_T)lnum, lines[i], 0, false) == FAIL) {
       api_set_error(err, kErrorTypeException, "Failed to insert line");
       goto end;
     }

     inserted_bytes += (bcount_t)strlen(lines[i]) + 1;

     extra++;
   }

   // Adjust marks. Invalidate any which lie in the
   // changed range, and move any in the remainder of the buffer.
   linenr_T adjust = end > start ? MAXLNUM : 0;
   mark_adjust_buf(buf, (linenr_T)start, (linenr_T)(end - 1), adjust, (linenr_T)extra,
                   true, kMarkAdjustApi, kExtmarkNOOP);

   extmark_splice(buf, (int)start - 1, 0, (int)(end - start), 0,
                  deleted_bytes, (int)new_len, 0, inserted_bytes,
                  kExtmarkUndo);

   changed_lines(buf, (linenr_T)start, 0, (linenr_T)end, (linenr_T)extra, true);

   FOR_ALL_TAB_WINDOWS(tp, win) {
     if (win->w_buffer == buf) {
       fix_cursor(win, (linenr_T)start, (linenr_T)end, (linenr_T)extra);
     }
   }
   end:;
 });
}

/// Sets (replaces) a range in the buffer
///
/// This is recommended over |nvim_buf_set_lines()| when only modifying parts of
/// a line, as extmarks will be preserved on non-modified parts of the touched
/// lines.
///
/// Indexing is zero-based. Row indices are end-inclusive, and column indices
/// are end-exclusive.
///
/// To insert text at a given `(row, column)` location, use `start_row = end_row
/// = row` and `start_col = end_col = col`. To delete the text in a range, use
/// `replacement = {}`.
///
/// @note Prefer |nvim_buf_set_lines()| (for performance) to add or delete entire lines.
/// @note Prefer |nvim_paste()| or |nvim_put()| to insert (instead of replace) text at cursor.
///
/// @param channel_id
/// @param buffer           Buffer id, or 0 for current buffer
/// @param start_row        First line index
/// @param start_col        Starting column (byte offset) on first line
/// @param end_row          Last line index, inclusive
/// @param end_col          Ending column (byte offset) on last line, exclusive
/// @param replacement      Array of lines to use as replacement
/// @param[out] err         Error details, if any
void nvim_buf_set_text(uint64_t channel_id, Buffer buffer, Integer start_row, Integer start_col,
                      Integer end_row, Integer end_col, ArrayOf(String) replacement, Arena *arena,
                      Error *err)
 FUNC_API_SINCE(7)
 FUNC_API_TEXTLOCK_ALLOW_CMDWIN
{
 MAXSIZE_TEMP_ARRAY(scratch, 1);
 if (replacement.size == 0) {
   ADD_C(scratch, STATIC_CSTR_AS_OBJ(""));
   replacement = scratch;
 }

 buf_T *buf = api_buf_ensure_loaded(buffer, err);
 if (!buf) {
   return;
 }

 bool oob = false;

 // check range is ordered and everything!
 // start_row, end_row within buffer len (except add text past the end?)
 start_row = normalize_index(buf, start_row, false, &oob);
 VALIDATE_RANGE((!oob), "start_row", {
   return;
 });

 end_row = normalize_index(buf, end_row, false, &oob);
 VALIDATE_RANGE((!oob), "end_row", {
   return;
 });

 // Another call to ml_get_buf() may free the lines, so we make copies
 char *str_at_start = ml_get_buf(buf, (linenr_T)start_row);
 colnr_T len_at_start = ml_get_buf_len(buf, (linenr_T)start_row);
 str_at_start = arena_memdupz(arena, str_at_start, (size_t)len_at_start);
 start_col = start_col < 0 ? len_at_start + start_col + 1 : start_col;
 VALIDATE_RANGE((start_col >= 0 && start_col <= len_at_start), "start_col", {
   return;
 });

 char *str_at_end = ml_get_buf(buf, (linenr_T)end_row);
 colnr_T len_at_end = ml_get_buf_len(buf, (linenr_T)end_row);
 str_at_end = arena_memdupz(arena, str_at_end, (size_t)len_at_end);
 end_col = end_col < 0 ? len_at_end + end_col + 1 : end_col;
 VALIDATE_RANGE((end_col >= 0 && end_col <= len_at_end), "end_col", {
   return;
 });

 VALIDATE((start_row <= end_row && !(end_row == start_row && start_col > end_col)),
          "%s", "'start' is higher than 'end'", {
   return;
 });

 bool disallow_nl = (channel_id != VIML_INTERNAL_CALL);
 if (!check_string_array(replacement, "replacement string", disallow_nl, err)) {
   return;
 }

 size_t new_len = replacement.size;

 bcount_t new_byte = 0;
 bcount_t old_byte = 0;

 // calculate byte size of old region before it gets modified/deleted
 if (start_row == end_row) {
   old_byte = (bcount_t)end_col - start_col;
 } else {
   old_byte += len_at_start - start_col;
   for (int64_t i = 1; i < end_row - start_row; i++) {
     int64_t lnum = start_row + i;
     old_byte += ml_get_buf_len(buf, (linenr_T)lnum) + 1;
   }
   old_byte += (bcount_t)end_col + 1;
 }

 String first_item = replacement.items[0].data.string;
 String last_item = replacement.items[replacement.size - 1].data.string;

 size_t firstlen = (size_t)start_col + first_item.size;
 size_t last_part_len = (size_t)len_at_end - (size_t)end_col;
 if (replacement.size == 1) {
   firstlen += last_part_len;
 }
 char *first = arena_allocz(arena, firstlen);
 char *last = NULL;
 memcpy(first, str_at_start, (size_t)start_col);
 memcpy(first + start_col, first_item.data, first_item.size);
 memchrsub(first + start_col, NUL, NL, first_item.size);
 if (replacement.size == 1) {
   memcpy(first + start_col + first_item.size, str_at_end + end_col, last_part_len);
 } else {
   last = arena_allocz(arena, last_item.size + last_part_len);
   memcpy(last, last_item.data, last_item.size);
   memchrsub(last, NUL, NL, last_item.size);
   memcpy(last + last_item.size, str_at_end + end_col, last_part_len);
 }

 char **lines = arena_alloc(arena, new_len * sizeof(char *), true);
 lines[0] = first;
 new_byte += (bcount_t)(first_item.size);
 for (size_t i = 1; i < new_len - 1; i++) {
   const String l = replacement.items[i].data.string;

   // Fill lines[i] with l's contents. Convert NULs to newlines as required by
   // NL-used-for-NUL.
   lines[i] = arena_memdupz(arena, l.data, l.size);
   memchrsub(lines[i], NUL, NL, l.size);
   new_byte += (bcount_t)(l.size) + 1;
 }
 if (replacement.size > 1) {
   lines[replacement.size - 1] = last;
   new_byte += (bcount_t)(last_item.size) + 1;
 }

 TRY_WRAP(err, {
   if (!MODIFIABLE(buf)) {
     api_set_error(err, kErrorTypeException, "Buffer is not 'modifiable'");
     goto end;
   }

   // Small note about undo states: unlike set_lines, we want to save the
   // undo state of one past the end_row, since end_row is inclusive.
   if (u_save_buf(buf, (linenr_T)start_row - 1, (linenr_T)end_row + 1) == FAIL) {
     api_set_error(err, kErrorTypeException, "Failed to save undo information");
     goto end;
   }

   ptrdiff_t extra = 0;  // lines added to text, can be negative
   size_t old_len = (size_t)(end_row - start_row + 1);

   // If the size of the range is reducing (ie, new_len < old_len) we
   // need to delete some old_len. We do this at the start, by
   // repeatedly deleting line "start".
   size_t to_delete = (new_len < old_len) ? old_len - new_len : 0;
   for (size_t i = 0; i < to_delete; i++) {
     if (ml_delete_buf(buf, (linenr_T)start_row, false) == FAIL) {
       api_set_error(err, kErrorTypeException, "Failed to delete line");
       goto end;
     }
   }

   if (to_delete > 0) {
     extra -= (ptrdiff_t)to_delete;
   }

   // For as long as possible, replace the existing old_len with the
   // new old_len. This is a more efficient operation, as it requires
   // less memory allocation and freeing.
   size_t to_replace = old_len < new_len ? old_len : new_len;
   for (size_t i = 0; i < to_replace; i++) {
     int64_t lnum = start_row + (int64_t)i;

     VALIDATE((lnum < MAXLNUM), "%s", "Index out of bounds", {
       goto end;
     });

     if (ml_replace_buf(buf, (linenr_T)lnum, lines[i], false, true) == FAIL) {
       api_set_error(err, kErrorTypeException, "Failed to replace line");
       goto end;
     }
   }

   // Now we may need to insert the remaining new old_len
   for (size_t i = to_replace; i < new_len; i++) {
     int64_t lnum = start_row + (int64_t)i - 1;

     VALIDATE((lnum < MAXLNUM), "%s", "Index out of bounds", {
       goto end;
     });

     if (ml_append_buf(buf, (linenr_T)lnum, lines[i], 0, false) == FAIL) {
       api_set_error(err, kErrorTypeException, "Failed to insert line");
       goto end;
     }

     extra++;
   }

   colnr_T col_extent = (colnr_T)(end_col
                                  - ((end_row == start_row) ? start_col : 0));

   // Adjust marks. Invalidate any which lie in the
   // changed range, and move any in the remainder of the buffer.
   // Do not adjust any cursors. need to use column-aware logic (below)
   linenr_T adjust = end_row >= start_row ? MAXLNUM : 0;
   mark_adjust_buf(buf, (linenr_T)start_row, (linenr_T)end_row - 1, adjust, (linenr_T)extra,
                   true, kMarkAdjustApi, kExtmarkNOOP);

   extmark_splice(buf, (int)start_row - 1, (colnr_T)start_col,
                  (int)(end_row - start_row), col_extent, old_byte,
                  (int)new_len - 1, (colnr_T)last_item.size, new_byte,
                  kExtmarkUndo);

   changed_lines(buf, (linenr_T)start_row, 0, (linenr_T)end_row + 1, (linenr_T)extra, true);

   FOR_ALL_TAB_WINDOWS(tp, win) {
     if (win->w_buffer == buf) {
       if (win->w_cursor.lnum >= start_row && win->w_cursor.lnum <= end_row) {
         fix_cursor_cols(win, (linenr_T)start_row, (colnr_T)start_col, (linenr_T)end_row,
                         (colnr_T)end_col, (linenr_T)new_len, (colnr_T)last_item.size);
       } else {
         fix_cursor(win, (linenr_T)start_row, (linenr_T)end_row, (linenr_T)extra);
       }
     }
   }
   end:;
 });
}

/// Gets a range from the buffer (may be partial lines, unlike |nvim_buf_get_lines()|).
///
/// Indexing is zero-based. Row indices are end-inclusive, and column indices
/// are end-exclusive.
///
/// Prefer |nvim_buf_get_lines()| when retrieving entire lines.
///
/// @param channel_id
/// @param buffer     Buffer id, or 0 for current buffer
/// @param start_row  First line index
/// @param start_col  Starting column (byte offset) on first line
/// @param end_row    Last line index, inclusive
/// @param end_col    Ending column (byte offset) on last line, exclusive
/// @param opts       Optional parameters. Currently unused.
/// @param[out] err   Error details, if any
/// @return Array of lines, or empty array for unloaded buffer.
ArrayOf(String) nvim_buf_get_text(uint64_t channel_id, Buffer buffer,
                                 Integer start_row, Integer start_col,
                                 Integer end_row, Integer end_col,
                                 Dict(empty) *opts,
                                 Arena *arena, lua_State *lstate, Error *err)
 FUNC_API_SINCE(9)
{
 Array rv = ARRAY_DICT_INIT;

 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return rv;
 }

 // return sentinel value if the buffer isn't loaded
 if (buf->b_ml.ml_mfp == NULL) {
   return rv;
 }

 bool oob = false;
 start_row = normalize_index(buf, start_row, false, &oob);
 end_row = normalize_index(buf, end_row, false, &oob);

 VALIDATE((!oob), "%s", "Index out of bounds", {
   return rv;
 });

 // nvim_buf_get_lines doesn't care if the start row is greater than the end
 // row (it will just return an empty array), but nvim_buf_get_text does in
 // order to maintain symmetry with nvim_buf_set_text.
 VALIDATE((start_row <= end_row), "%s", "'start' is higher than 'end'", {
   return rv;
 });

 bool replace_nl = (channel_id != VIML_INTERNAL_CALL);

 size_t size = (size_t)(end_row - start_row) + 1;

 init_line_array(lstate, &rv, size, arena);

 if (start_row == end_row) {
   String line = buf_get_text(buf, start_row, start_col, end_col, err);
   if (ERROR_SET(err)) {
     goto end;
   }
   push_linestr(lstate, &rv, line.data, line.size, 0, replace_nl, arena);
   return rv;
 }

 String str = buf_get_text(buf, start_row, start_col, MAXCOL - 1, err);
 if (ERROR_SET(err)) {
   goto end;
 }

 push_linestr(lstate, &rv, str.data, str.size, 0, replace_nl, arena);

 if (size > 2) {
   buf_collect_lines(buf, size - 2, (linenr_T)start_row + 1, 1, replace_nl, &rv, lstate, arena);
 }

 str = buf_get_text(buf, end_row, 0, end_col, err);
 if (ERROR_SET(err)) {
   goto end;
 }

 push_linestr(lstate, &rv, str.data, str.size, (int)(size - 1), replace_nl, arena);

end:
 if (ERROR_SET(err)) {
   return (Array)ARRAY_DICT_INIT;
 }

 return rv;
}

/// Returns the byte offset of a line (0-indexed). |api-indexing|
///
/// Line 1 (index=0) has offset 0. UTF-8 bytes are counted. EOL is one byte.
/// 'fileformat' and 'fileencoding' are ignored. The line index just after the
/// last line gives the total byte-count of the buffer. A final EOL byte is
/// counted if it would be written, see 'eol'.
///
/// Unlike |line2byte()|, throws error for out-of-bounds indexing.
/// Returns -1 for unloaded buffer.
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param index      Line index
/// @param[out] err   Error details, if any
/// @return Integer byte offset, or -1 for unloaded buffer.
Integer nvim_buf_get_offset(Buffer buffer, Integer index, Error *err)
 FUNC_API_SINCE(5)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);
 if (!buf) {
   return 0;
 }

 // return sentinel value if the buffer isn't loaded
 if (buf->b_ml.ml_mfp == NULL) {
   return -1;
 }

 VALIDATE((index >= 0 && index <= buf->b_ml.ml_line_count), "%s", "Index out of bounds", {
   return 0;
 });

 return ml_find_line_or_offset(buf, (int)index + 1, NULL, true);
}

/// Gets a buffer-scoped (b:) variable.
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param name       Variable name
/// @param[out] err   Error details, if any
/// @return Variable value
Object nvim_buf_get_var(Buffer buffer, String name, Arena *arena, Error *err)
 FUNC_API_SINCE(1)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return (Object)OBJECT_INIT;
 }

 return dict_get_value(buf->b_vars, name, arena, err);
}

/// Gets a changed tick of a buffer
///
/// @param[in]  buffer  Buffer id, or 0 for current buffer
/// @param[out] err     Error details, if any
///
/// @return `b:changedtick` value.
Integer nvim_buf_get_changedtick(Buffer buffer, Error *err)
 FUNC_API_SINCE(2)
{
 const buf_T *const buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return -1;
 }

 return buf_get_changedtick(buf);
}

/// Gets a list of buffer-local |mapping| definitions.
///
/// @param  buffer     Buffer id, or 0 for current buffer
/// @param  mode       Mode short-name ("n", "i", "v", ...)
/// @param[out]  err   Error details, if any
/// @returns Array of |maparg()|-like dictionaries describing mappings.
///          The "buffer" key holds the associated buffer id.
ArrayOf(DictAs(get_keymap)) nvim_buf_get_keymap(Buffer buffer, String mode, Arena *arena,
                                               Error *err)
 FUNC_API_SINCE(3)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return (Array)ARRAY_DICT_INIT;
 }

 return keymap_array(mode, buf, arena);
}

/// Sets a buffer-local |mapping| for the given mode.
///
/// @see |nvim_set_keymap()|
///
/// @param  buffer  Buffer id, or 0 for current buffer
void nvim_buf_set_keymap(uint64_t channel_id, Buffer buffer, String mode, String lhs, String rhs,
                        Dict(keymap) *opts, Error *err)
 FUNC_API_SINCE(6)
{
 modify_keymap(channel_id, buffer, false, mode, lhs, rhs, opts, err);
}

/// Unmaps a buffer-local |mapping| for the given mode.
///
/// @see |nvim_del_keymap()|
///
/// @param  buffer  Buffer id, or 0 for current buffer
void nvim_buf_del_keymap(uint64_t channel_id, Buffer buffer, String mode, String lhs, Error *err)
 FUNC_API_SINCE(6)
{
 String rhs = { .data = "", .size = 0 };
 modify_keymap(channel_id, buffer, true, mode, lhs, rhs, NULL, err);
}

/// Sets a buffer-scoped (b:) variable
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param name       Variable name
/// @param value      Variable value
/// @param[out] err   Error details, if any
void nvim_buf_set_var(Buffer buffer, String name, Object value, Error *err)
 FUNC_API_SINCE(1)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return;
 }

 dict_set_var(buf->b_vars, name, value, false, false, NULL, err);
}

/// Removes a buffer-scoped (b:) variable
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param name       Variable name
/// @param[out] err   Error details, if any
void nvim_buf_del_var(Buffer buffer, String name, Error *err)
 FUNC_API_SINCE(1)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return;
 }

 dict_set_var(buf->b_vars, name, NIL, true, false, NULL, err);
}

/// Gets the full file name for the buffer
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param[out] err   Error details, if any
/// @return Buffer name
String nvim_buf_get_name(Buffer buffer, Error *err)
 FUNC_API_SINCE(1)
{
 String rv = STRING_INIT;
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf || buf->b_ffname == NULL) {
   return rv;
 }

 return cstr_as_string(buf->b_ffname);
}

/// Sets the full file name for a buffer, like |:file_f|
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param name       Buffer name
/// @param[out] err   Error details, if any
void nvim_buf_set_name(Buffer buffer, String name, Error *err)
 FUNC_API_SINCE(1)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return;
 }

 int ren_ret = OK;
 TRY_WRAP(err, {
   const bool is_curbuf = buf == curbuf;
   const int save_acd = p_acd;
   if (!is_curbuf) {
     // Temporarily disable 'autochdir' when setting file name for another buffer.
     p_acd = false;
   }

   // Using aucmd_*: autocommands will be executed by rename_buffer
   aco_save_T aco;
   aucmd_prepbuf(&aco, buf);
   ren_ret = rename_buffer(name.data);
   aucmd_restbuf(&aco);

   if (!is_curbuf) {
     p_acd = save_acd;
   }
 });

 if (ERROR_SET(err)) {
   return;
 }

 if (ren_ret == FAIL) {
   api_set_error(err, kErrorTypeException, "Failed to rename buffer");
 }
}

/// Checks if a buffer is valid and loaded. See |api-buffer| for more info
/// about unloaded buffers.
///
/// @param buffer Buffer id, or 0 for current buffer
/// @return true if the buffer is valid and loaded, false otherwise.
Boolean nvim_buf_is_loaded(Buffer buffer)
 FUNC_API_SINCE(5)
{
 Error stub = ERROR_INIT;
 buf_T *buf = find_buffer_by_handle(buffer, &stub);
 api_clear_error(&stub);
 return buf && buf->b_ml.ml_mfp != NULL;
}

/// Deletes a buffer and its metadata (like |:bwipeout|).
///
/// To get |:bdelete| behavior, reset 'buflisted' and pass `unload=true`:
/// ```lua
/// vim.bo.buflisted = false
/// vim.api.nvim_buf_delete(0, { unload = true })
/// ```
///
/// @param buffer Buffer id, or 0 for current buffer
/// @param opts  Optional parameters. Keys:
///          - force:  Force deletion, ignore unsaved changes.
///          - unload: Unloaded only (|:bunload|), do not delete.
void nvim_buf_delete(Buffer buffer, Dict(buf_delete) *opts, Error *err)
 FUNC_API_SINCE(7)
 FUNC_API_TEXTLOCK
{
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (ERROR_SET(err)) {
   return;
 }

 bool force = opts->force;

 bool unload = opts->unload;

 int result = do_buffer(unload ? DOBUF_UNLOAD : DOBUF_WIPE,
                        DOBUF_FIRST,
                        FORWARD,
                        buf->handle,
                        force);

 if (result == FAIL) {
   api_set_error(err, kErrorTypeException, "Failed to unload buffer.");
   return;
 }
}

/// Checks if a buffer is valid.
///
/// @note Even if a buffer is valid it may have been unloaded. See |api-buffer|
/// for more info about unloaded buffers.
///
/// @param buffer Buffer id, or 0 for current buffer
/// @return true if the buffer is valid, false otherwise.
Boolean nvim_buf_is_valid(Buffer buffer)
 FUNC_API_SINCE(1)
{
 Error stub = ERROR_INIT;
 Boolean ret = find_buffer_by_handle(buffer, &stub) != NULL;
 api_clear_error(&stub);
 return ret;
}

/// Deletes a named mark in the buffer. See |mark-motions|.
///
/// @note only deletes marks set in the buffer, if the mark is not set
/// in the buffer it will return false.
/// @param buffer     Buffer to set the mark on
/// @param name       Mark name
/// @return true if the mark was deleted, else false.
/// @see |nvim_buf_set_mark()|
/// @see |nvim_del_mark()|
Boolean nvim_buf_del_mark(Buffer buffer, String name, Error *err)
 FUNC_API_SINCE(8)
{
 bool res = false;
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return res;
 }

 VALIDATE_S((name.size == 1), "mark name (must be a single char)", name.data, {
   return res;
 });

 fmark_T *fm = mark_get(buf, curwin, NULL, kMarkAllNoResolve, *name.data);

 // fm is NULL when there's no mark with the given name
 VALIDATE_S((fm != NULL), "mark name", name.data, {
   return res;
 });

 // mark.lnum is 0 when the mark is not valid in the buffer, or is not set.
 if (fm->mark.lnum != 0 && fm->fnum == buf->handle) {
   // since the mark belongs to the buffer delete it.
   res = set_mark(buf, name, 0, 0, err);
 }

 return res;
}

/// Sets a named mark in the given buffer, all marks are allowed
/// file/uppercase, visual, last change, etc. See |mark-motions|.
///
/// Marks are (1,0)-indexed. |api-indexing|
///
/// @note Passing 0 as line deletes the mark
///
/// @param buffer     Buffer to set the mark on
/// @param name       Mark name
/// @param line       Line number
/// @param col        Column/row number
/// @param opts       Optional parameters. Reserved for future use.
/// @return true if the mark was set, else false.
/// @see |nvim_buf_del_mark()|
/// @see |nvim_buf_get_mark()|
Boolean nvim_buf_set_mark(Buffer buffer, String name, Integer line, Integer col, Dict(empty) *opts,
                         Error *err)
 FUNC_API_SINCE(8)
{
 bool res = false;
 buf_T *buf = api_buf_ensure_loaded(buffer, err);

 if (!buf) {
   return res;
 }

 VALIDATE_S((name.size == 1), "mark name (must be a single char)", name.data, {
   return res;
 });

 res = set_mark(buf, name, line, col, err);

 return res;
}

/// Returns a `(row,col)` tuple representing the position of the named mark.
/// "End of line" column position is returned as |v:maxcol| (big number).
/// See |mark-motions|.
///
/// Marks are (1,0)-indexed. |api-indexing|
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param name       Mark name
/// @param[out] err   Error details, if any
/// @return (row, col) tuple, (0, 0) if the mark is not set, or is an
/// uppercase/file mark set in another buffer.
/// @see |nvim_buf_set_mark()|
/// @see |nvim_buf_del_mark()|
ArrayOf(Integer, 2) nvim_buf_get_mark(Buffer buffer, String name, Arena *arena, Error *err)
 FUNC_API_SINCE(1)
{
 Array rv = ARRAY_DICT_INIT;
 buf_T *buf = find_buffer_by_handle(buffer, err);

 if (!buf) {
   return rv;
 }

 VALIDATE_S((name.size == 1), "mark name (must be a single char)", name.data, {
   return rv;
 });

 fmark_T *fm;
 pos_T pos;
 char mark = *name.data;

 fm = mark_get(buf, curwin, NULL, kMarkAllNoResolve, mark);
 VALIDATE_S((fm != NULL), "mark name", name.data, {
   return rv;
 });
 // (0, 0) uppercase/file mark set in another buffer.
 if (fm->fnum != buf->handle) {
   pos.lnum = 0;
   pos.col = 0;
 } else {
   pos = fm->mark;
 }

 rv = arena_array(arena, 2);
 ADD_C(rv, INTEGER_OBJ(pos.lnum));
 ADD_C(rv, INTEGER_OBJ(pos.col));

 return rv;
}

/// Call a function with buffer as temporary current buffer.
///
/// This temporarily switches current buffer to "buffer".
/// If the current window already shows "buffer", the window is not switched.
/// If a window inside the current tabpage (including a float) already shows the
/// buffer, then one of those windows will be set as current window temporarily.
/// Otherwise a temporary scratch window (called the "autocmd window" for
/// historical reasons) will be used.
///
/// This is useful e.g. to call Vimscript functions that only work with the
/// current buffer/window currently, like `jobstart(…, {'term': v:true})`.
///
/// @param buffer     Buffer id, or 0 for current buffer
/// @param fun        Function to call inside the buffer (currently Lua callable
///                   only)
/// @param[out] err   Error details, if any
/// @return           Return value of function.
Object nvim_buf_call(Buffer buffer, LuaRef fun, Error *err)
 FUNC_API_SINCE(7)
 FUNC_API_LUA_ONLY
{
 buf_T *buf = find_buffer_by_handle(buffer, err);
 if (!buf) {
   return NIL;
 }

 Object res = OBJECT_INIT;
 TRY_WRAP(err, {
   aco_save_T aco;
   aucmd_prepbuf(&aco, buf);

   Array args = ARRAY_DICT_INIT;
   res = nlua_call_ref(fun, NULL, args, kRetLuaref, NULL, err);

   aucmd_restbuf(&aco);
 });

 return res;
}

/// @nodoc
Dict nvim__buf_stats(Buffer buffer, Arena *arena, Error *err)
{
 buf_T *buf = find_buffer_by_handle(buffer, err);
 if (!buf) {
   return (Dict)ARRAY_DICT_INIT;
 }

 Dict rv = arena_dict(arena, 7);
 // Number of times the cached line was flushed.
 // This should generally not increase while editing the same
 // line in the same mode.
 PUT_C(rv, "flush_count", INTEGER_OBJ(buf->flush_count));
 // lnum of current line
 PUT_C(rv, "current_lnum", INTEGER_OBJ(buf->b_ml.ml_line_lnum));
 // whether the line has unflushed changes.
 PUT_C(rv, "line_dirty", BOOLEAN_OBJ(buf->b_ml.ml_flags & ML_LINE_DIRTY));
 // NB: this should be zero at any time API functions are called,
 // this exists to debug issues
 PUT_C(rv, "dirty_bytes", INTEGER_OBJ((Integer)buf->deleted_bytes));
 PUT_C(rv, "dirty_bytes2", INTEGER_OBJ((Integer)buf->deleted_bytes2));
 PUT_C(rv, "virt_blocks", INTEGER_OBJ((Integer)buf_meta_total(buf, kMTMetaLines)));

 u_header_T *uhp = NULL;
 if (buf->b_u_curhead != NULL) {
   uhp = buf->b_u_curhead;
 } else if (buf->b_u_newhead) {
   uhp = buf->b_u_newhead;
 }
 if (uhp) {
   PUT_C(rv, "uhp_extmark_size", INTEGER_OBJ((Integer)kv_size(uhp->uh_extmark)));
 }

 return rv;
}

// Check if deleting lines made the cursor position invalid.
// Changed lines from `lo` to `hi`; added `extra` lines (negative if deleted).
static void fix_cursor(win_T *win, linenr_T lo, linenr_T hi, linenr_T extra)
{
 if (win->w_cursor.lnum >= lo) {
   // Adjust cursor position if it's in/after the changed lines.
   if (win->w_cursor.lnum >= hi) {
     win->w_cursor.lnum += extra;
   } else if (extra < 0) {
     check_cursor_lnum(win);
   }
   check_cursor_col(win);
   changed_cline_bef_curs(win);
   win->w_valid &= ~(VALID_BOTLINE_AP);
   update_topline(win);
 } else {
   invalidate_botline_win(win);
 }
}

/// Fix cursor position after replacing text
/// between (start_row, start_col) and (end_row, end_col).
///
/// win->w_cursor.lnum is assumed to be >= start_row and <= end_row.
static void fix_cursor_cols(win_T *win, linenr_T start_row, colnr_T start_col, linenr_T end_row,
                           colnr_T end_col, linenr_T new_rows, colnr_T new_cols_at_end_row)
{
 colnr_T mode_col_adj = win == curwin && (State & MODE_INSERT) ? 0 : 1;

 colnr_T end_row_change_start = new_rows == 1 ? start_col : 0;
 colnr_T end_row_change_end = end_row_change_start + new_cols_at_end_row;

 // check if cursor is after replaced range or not
 if (win->w_cursor.lnum == end_row && win->w_cursor.col + mode_col_adj > end_col) {
   // if cursor is after replaced range, it's shifted
   // to keep it's position the same, relative to end_col

   linenr_T old_rows = end_row - start_row + 1;
   win->w_cursor.lnum += new_rows - old_rows;
   win->w_cursor.col += end_row_change_end - end_col;
 } else {
   // if cursor is inside replaced range
   // and the new range got smaller,
   // it's shifted to keep it inside the new range
   //
   // if cursor is before range or range did not
   // got smaller, position is not changed

   colnr_T old_coladd = win->w_cursor.coladd;

   // it's easier to work with a single value here.
   // col and coladd are fixed by a later call
   // to check_cursor_col when necessary
   win->w_cursor.col += win->w_cursor.coladd;
   win->w_cursor.coladd = 0;

   linenr_T new_end_row = start_row + new_rows - 1;

   // make sure cursor row is in the new row range
   if (win->w_cursor.lnum > new_end_row) {
     win->w_cursor.lnum = new_end_row;

     // don't simply move cursor up, but to the end
     // of new_end_row, if it's not at or after
     // it already (in case virtualedit is active)
     // column might be additionally adjusted below
     // to keep it inside col range if needed
     colnr_T len = ml_get_buf_len(win->w_buffer, new_end_row);
     if (win->w_cursor.col < len) {
       win->w_cursor.col = len;
     }
   }

   // if cursor is at the last row and
   // it wasn't after eol before, move it exactly
   // to end_row_change_end
   if (win->w_cursor.lnum == new_end_row
       && win->w_cursor.col > end_row_change_end && old_coladd == 0) {
     win->w_cursor.col = end_row_change_end;

     // make sure cursor is inside range, not after it,
     // except when doing so would move it before new range
     if (win->w_cursor.col - mode_col_adj >= end_row_change_start) {
       win->w_cursor.col -= mode_col_adj;
     }
   }
 }

 check_cursor_col(win);
 changed_cline_bef_curs(win);
 invalidate_botline_win(win);
}

/// Initialise a string array either:
/// - on the Lua stack (as a table) (if lstate is not NULL)
/// - as an API array object (if lstate is NULL).
///
/// @param lstate  Lua state. When NULL the Array is initialized instead.
/// @param a       Array to initialize
/// @param size    Size of array
static inline void init_line_array(lua_State *lstate, Array *a, size_t size, Arena *arena)
{
 if (lstate) {
   lua_createtable(lstate, (int)size, 0);
 } else {
   *a = arena_array(arena, size);
 }
}

/// Push a string onto either the Lua stack (as a table element) or an API array object.
///
/// For Lua, a table of the correct size must be created first.
/// API array objects must be pre allocated.
///
/// @param lstate      Lua state. When NULL the Array is pushed to instead.
/// @param a           Array to push onto when not using Lua
/// @param s           String to push
/// @param len         Size of string
/// @param idx         0-based index to place s (only used for Lua)
/// @param replace_nl  Replace newlines ('\n') with null (NUL)
static void push_linestr(lua_State *lstate, Array *a, const char *s, size_t len, int idx,
                        bool replace_nl, Arena *arena)
{
 if (lstate) {
   // Vim represents NULs as NLs
   if (s && replace_nl && strchr(s, '\n')) {
     // TODO(bfredl): could manage scratch space in the arena, for the NUL case
     char *tmp = xmemdupz(s, len);
     strchrsub(tmp, '\n', NUL);
     lua_pushlstring(lstate, tmp, len);
     xfree(tmp);
   } else {
     lua_pushlstring(lstate, s, len);
   }
   lua_rawseti(lstate, -2, idx + 1);
 } else {
   String str = STRING_INIT;
   if (len > 0) {
     str = CBUF_TO_ARENA_STR(arena, s, len);
     if (replace_nl) {
       // Vim represents NULs as NLs, but this may confuse clients.
       strchrsub(str.data, '\n', NUL);
     }
   }

   ADD_C(*a, STRING_OBJ(str));
 }
}

/// Collects `n` buffer lines into array `l` and/or lua_State `lstate`, optionally replacing
/// newlines with NUL.
///
/// @param buf Buffer to get lines from
/// @param n Number of lines to collect
/// @param replace_nl Replace newlines ("\n") with NUL
/// @param start Line number to start from
/// @param start_idx First index to push to (only used for Lua)
/// @param[out] l If not NULL, Lines are copied here
/// @param[out] lstate If not NULL, Lines are pushed into a table onto the stack
/// @param err[out] Error, if any
/// @return true unless `err` was set
void buf_collect_lines(buf_T *buf, size_t n, linenr_T start, int start_idx, bool replace_nl,
                      Array *l, lua_State *lstate, Arena *arena)
{
 for (size_t i = 0; i < n; i++) {
   linenr_T lnum = start + (linenr_T)i;
   char *bufstr = ml_get_buf(buf, lnum);
   size_t bufstrlen = (size_t)ml_get_buf_len(buf, lnum);
   push_linestr(lstate, l, bufstr, bufstrlen, start_idx + (int)i, replace_nl, arena);
 }
}