Merge upstream, and index pressure advance by extruder variant

Upstream moved pressure advance onto the extruder variant: enable_pressure_advance,
pressure_advance and the four adaptive keys joined filament_options_with_variant,
the repeated inline blocks in set_extruder() became a helper, and the lookups
moved from the filament id to get_filament_config_index().

All three conflicts were the same collision, because this branch had modified two
of those same inline blocks to pass a tool qualifier so each carriage is addressed
explicitly in parallel modes. Taking either side whole would have lost something:
upstream's drops the qualifier and leaves parallel carriages with no pressure
advance, ours drops the per-variant indexing and reads the wrong column on a
multi-variant printer. The helper now takes an optional tool, defaulting to -1,
which omits the qualifier. imex_pem_tool_for() already returns -1 off IMEX and in
primary mode, so non-IMEX output is unchanged, and the three call sites that never
passed a tool keep upstream's behavior exactly.

The third conflict was two test cases appended at the same place. Both are kept.

Separately, one defect that merged cleanly and so was not flagged: the loop that
emits pressure advance for secondary carriages at the start of a print still
bounded and indexed those vectors with a raw filament id. They are variant
expanded now, so their length is columns rather than filament slots -- the value
read was the wrong column, and the bound no longer sat in slot space, letting an
out-of-slot filament through. It now bounds on filament_diameter and translates
with get_filament_config_index(), which is what the sibling second-layer
temperature loop already does.

Verified: both changed translation units compile clean under -Werror. The merge
was resolved independently twice and the two resolutions agree on every line of
code. Not yet run: the Release test suite and a parallel-mode slice sweep.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Clifford Garwood
2026-09-29 21:43:47 -04:00
co-authored by Claude Opus 5
109 changed files with 3823 additions and 1295 deletions
+88 -101
View File
@@ -1499,12 +1499,7 @@ static std::vector<Vec2d> get_path_of_change_filament(const Print& print)
check_add_eol(toolchange_gcode_str);
// SoftFever: set new PA for new filament
if (gcodegen.config().enable_pressure_advance.get_at(new_filament_id)) {
gcode += gcodegen.writer().set_pressure_advance(gcodegen.config().pressure_advance.get_at(new_filament_id));
// Orca: Adaptive PA
// Reset Adaptive PA processor last PA value
gcodegen.m_pa_processor->resetPreviousPA(gcodegen.config().pressure_advance.get_at(new_filament_id));
}
gcode += gcodegen.set_filament_pressure_advance(new_filament_id);
// A phony move to the end position at the wipe tower.
gcodegen.writer().travel_to_xy((end_pos + plate_origin_2d).cast<double>());
@@ -1799,12 +1794,8 @@ static std::vector<Vec2d> get_path_of_change_filament(const Print& print)
check_add_eol(toolchange_gcode_str);
// SoftFever: set new PA for new filament
if (new_extruder_id != -1 && gcodegen.config().enable_pressure_advance.get_at(new_extruder_id)) {
gcode += gcodegen.writer().set_pressure_advance(gcodegen.config().pressure_advance.get_at(new_extruder_id));
// Orca: Adaptive PA
// Reset Adaptive PA processor last PA value
gcodegen.m_pa_processor->resetPreviousPA(gcodegen.config().pressure_advance.get_at(new_extruder_id));
}
if (new_extruder_id != -1)
gcode += gcodegen.set_filament_pressure_advance(new_extruder_id);
// A phony move to the end position at the wipe tower.
gcodegen.writer().travel_to_xy((end_pos + plate_origin_2d).cast<double>());
@@ -3237,7 +3228,7 @@ void GCode::_do_export(Print& print, GCodeOutputStream &file, ThumbnailsGenerato
m_cooling_buffer->set_current_extruder(initial_extruder_id, extruder_id);
// Orca: Initialise AdaptivePA processor filter
m_pa_processor = std::make_unique<AdaptivePAProcessor>(*this, tool_ordering.all_extruders());
m_pa_processor = std::make_unique<AdaptivePAProcessor>(*this);
// Update output variables after the extruders were initialized.
m_placeholder_parser_integration.init(m_writer);
@@ -3948,12 +3939,7 @@ void GCode::_do_export(Print& print, GCodeOutputStream &file, ThumbnailsGenerato
}
}
// Orca: add missing PA settings for initial filament
if (m_config.enable_pressure_advance.get_at(initial_non_support_extruder_id)) {
file.write(m_writer.set_pressure_advance(m_config.pressure_advance.get_at(initial_non_support_extruder_id)));
// Orca: Adaptive PA
// Reset Adaptive PA processor last PA value
m_pa_processor->resetPreviousPA(m_config.pressure_advance.get_at(initial_non_support_extruder_id));
}
file.write(set_filament_pressure_advance(initial_non_support_extruder_id));
}
//flush FanMover buffer to avoid modifying the start gcode if it's manual.
@@ -4040,9 +4026,11 @@ void GCode::_do_export(Print& print, GCodeOutputStream &file, ThumbnailsGenerato
(int) initial_extruder_id < (int) m_config.physical_extruder_map.values.size())
? m_config.physical_extruder_map.values[(int) initial_extruder_id]
: -1;
const int num_pa_filament_columns = std::min(
(int) print.config().enable_pressure_advance.values.size(),
(int) print.config().pressure_advance.values.size());
// enable_pressure_advance and pressure_advance are variant-expanded, so their
// length is columns, not filament slots. Bound in slot space -- filament_diameter
// is one entry per slot -- exactly as the second-layer temperature loop does, then
// translate the slot to its column with get_filament_config_index().
const int num_pa_filament_slots = (int) print.config().filament_diameter.values.size();
for (int tool_idx : get_imex_active_tools(print)) {
// Unlike the second-layer temperature loop, the primary is skipped here:
// set_extruder() above already emitted its PA with the pem tool qualifier.
@@ -4054,10 +4042,11 @@ void GCode::_do_export(Print& print, GCodeOutputStream &file, ThumbnailsGenerato
// diverge when the printer has more nozzles than the project has filaments, and
// get_at() would clamp a past-the-end index to filament 0 and pin its PA onto a
// secondary carriage. Same guard the second-layer temperature loop uses.
if (logical < 0 || logical >= num_pa_filament_columns) continue;
if (!print.config().enable_pressure_advance.get_at(logical)) continue;
if (logical < 0 || logical >= num_pa_filament_slots) continue;
const size_t pa_fi = get_filament_config_index(logical);
if (!print.config().enable_pressure_advance.get_at(pa_fi)) continue;
file.write(m_writer.set_pressure_advance(
print.config().pressure_advance.get_at(logical),
print.config().pressure_advance.get_at(pa_fi),
tool_idx));
}
}
@@ -4379,6 +4368,28 @@ void GCode::_do_export(Print& print, GCodeOutputStream &file, ThumbnailsGenerato
print.throw_if_canceled();
// Some firmwares only scan the last N lines for the time estimate, so the stats are written
// after the config block.
if (!is_bbl_printers && !skip_config_block) {
file.write("; CONFIG_BLOCK_START\n");
std::string full_config;
append_full_config(print, full_config);
if (!full_config.empty())
file.write(full_config);
// SoftFever: write compatiple info
int first_layer_bed_temperature = get_bed_temperature(0, true, print.config().curr_bed_type);
file.write_format("; first_layer_bed_temperature = %d\n", first_layer_bed_temperature);
file.write_format("; bed_shape = %s\n", print.full_print_config().opt_serialize("printable_area").c_str());
file.write_format("; first_layer_temperature = %d\n", print.config().nozzle_temperature_initial_layer.get_at(0));
file.write_format("; first_layer_height = %.3f\n", print.config().initial_layer_print_height.value);
//SF TODO
// file.write_format("; variable_layer_height = %d\n", print.ad.adaptive_layer_height ? 1 : 0);
file.write("; CONFIG_BLOCK_END\n\n");
}
// Get filament stats.
file.write(DoExport::update_print_stats_and_format_filament_stats(
// Const inputs
@@ -4405,26 +4416,6 @@ void GCode::_do_export(Print& print, GCodeOutputStream &file, ThumbnailsGenerato
GCodeProcessor::ETags::Estimated_Printing_Time_Placeholder)
.c_str());
file.write("\n");
if (!skip_config_block) {
file.write("; CONFIG_BLOCK_START\n");
std::string full_config;
append_full_config(print, full_config);
if (!full_config.empty())
file.write(full_config);
// SoftFever: write compatiple info
int first_layer_bed_temperature = get_bed_temperature(0, true, print.config().curr_bed_type);
file.write_format("; first_layer_bed_temperature = %d\n", first_layer_bed_temperature);
file.write_format("; bed_shape = %s\n", print.full_print_config().opt_serialize("printable_area").c_str());
file.write_format("; first_layer_temperature = %d\n", print.config().nozzle_temperature_initial_layer.get_at(0));
file.write_format("; first_layer_height = %.3f\n", print.config().initial_layer_print_height.value);
//SF TODO
// file.write_format("; variable_layer_height = %d\n", print.ad.adaptive_layer_height ? 1 : 0);
file.write("; CONFIG_BLOCK_END\n\n");
} // !skip_config_block
}
file.write("\n");
@@ -8647,12 +8638,16 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
// Orca: Dynamic PA
// If adaptive PA is enabled, by default evaluate PA on all extrusion moves
// The PA settings and the PA_CHANGE tags use the filament config index of the filament's extruder variant
const unsigned int pa_config_index = (unsigned int) get_filament_config_index(m_writer.filament()->id());
const bool adaptive_pa = m_config.adaptive_pressure_advance.get_at(pa_config_index) && m_config.enable_pressure_advance.get_at(pa_config_index);
const bool adaptive_pa_overhangs = adaptive_pa && m_config.adaptive_pressure_advance_overhangs.get_at(pa_config_index);
bool is_pa_calib = m_curr_print->calib_mode() == CalibMode::Calib_PA_Line ||
m_curr_print->calib_mode() == CalibMode::Calib_PA_Pattern ||
m_curr_print->calib_mode() == CalibMode::Calib_PA_Tower;
bool evaluate_adaptive_pa = false;
bool role_change = (m_last_extrusion_role != path.role());
if (!is_pa_calib && FILAMENT_CONFIG(adaptive_pressure_advance) && FILAMENT_CONFIG(enable_pressure_advance)) {
if (!is_pa_calib && adaptive_pa) {
evaluate_adaptive_pa = true;
// If we have already emmited a PA change because the m_multi_flow_segment_path_pa_set is set
// skip re-issuing the PA change tag.
@@ -8735,7 +8730,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
// Post processor flag generation code segment when option to emit only at role changes is enabled
// Variables published to the post processor:
// 1) Tag to trigger a PA evaluation (because a role change was identified and the user has requested dynamic PA adjustments)
// 2) Current extruder ID (to identify the PA model for the currently used extruder)
// 2) Current filament config index (to identify the PA model for the currently used filament and its extruder variant)
// 3) mm3_per_mm value (to then multiply by the final model print speed after slowdown for cooling is applied)
// 4) the current acceleration (to pass to the model for evaluation)
// 5) whether this is an external perimeter (for future use)
@@ -8746,7 +8741,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
if (m_multi_flow_segment_path_average_mm3_per_mm > 0) {
sprintf(buf, ";%sT%u MM3MM:%g ACCEL:%u BR:%d RC:%d OV:%d\n",
GCodeProcessor::reserved_tag(GCodeProcessor::ETags::PA_Change).c_str(),
m_writer.filament()->id(),
pa_config_index,
m_multi_flow_segment_path_average_mm3_per_mm,
acceleration_i,
((path.role() == erBridgeInfill) ||(path.role() == erOverhangPerimeter)),
@@ -8759,7 +8754,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
// to issue a zero flow PA change command for this
sprintf(buf, ";%sT%u MM3MM:%g ACCEL:%u BR:%d RC:%d OV:%d\n",
GCodeProcessor::reserved_tag(GCodeProcessor::ETags::PA_Change).c_str(),
m_writer.filament()->id(),
pa_config_index,
_mm3_per_mm,
acceleration_i,
((path.role() == erBridgeInfill) ||(path.role() == erOverhangPerimeter)),
@@ -8867,11 +8862,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
// or a flow change, so emit the flag to evaluate PA for the upcomming extrusion
// Emit tag before new speed is set so the post processor reads the next speed immediately and uses it.
// Dont emit tag if it has just already been emitted from a role change above
if(_mm3_per_mm >0 &&
FILAMENT_CONFIG(adaptive_pressure_advance) &&
FILAMENT_CONFIG(enable_pressure_advance) &&
FILAMENT_CONFIG(adaptive_pressure_advance_overhangs) &&
!evaluate_adaptive_pa){
if(_mm3_per_mm >0 && adaptive_pa_overhangs && !evaluate_adaptive_pa){
if(writer().get_current_speed() > F){ // Ramping down speed - use overhang logic where the minimum speed is used between current and upcoming extrusion
if(m_config.gcode_comments){
sprintf(buf, "; Ramp down-non-variable\n");
@@ -8879,7 +8870,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
}
sprintf(buf, ";%sT%u MM3MM:%g ACCEL:%u BR:%d RC:%d OV:%d\n",
GCodeProcessor::reserved_tag(GCodeProcessor::ETags::PA_Change).c_str(),
m_writer.filament()->id(),
pa_config_index,
_mm3_per_mm,
acceleration_i,
((path.role() == erBridgeInfill) ||(path.role() == erOverhangPerimeter)),
@@ -8894,7 +8885,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
}
sprintf(buf, ";%sT%u MM3MM:%g ACCEL:%u BR:%d RC:%d OV:%d\n",
GCodeProcessor::reserved_tag(GCodeProcessor::ETags::PA_Change).c_str(),
m_writer.filament()->id(),
pa_config_index,
_mm3_per_mm,
acceleration_i,
((path.role() == erBridgeInfill) ||(path.role() == erOverhangPerimeter)),
@@ -9100,10 +9091,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
// ORCA: Adaptive PA code segment when adjusting PA within the same feature
// There is a speed change or flow change so emit the flag to evaluate PA for the upcomming extrusion
// Emit tag before new speed is set so the post processor reads the next speed immediately and uses it.
if(_mm3_per_mm >0 &&
EXTRUDER_CONFIG(adaptive_pressure_advance) &&
EXTRUDER_CONFIG(enable_pressure_advance) &&
EXTRUDER_CONFIG(adaptive_pressure_advance_overhangs) ){
if(_mm3_per_mm >0 && adaptive_pa_overhangs){
if(last_set_speed > new_speed){ // Ramping down speed - use overhang logic where the minimum speed is used between current and upcoming extrusion
if(m_config.gcode_comments) {
sprintf(buf, "; Ramp up-variable\n");
@@ -9111,7 +9099,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
}
sprintf(buf, ";%sT%u MM3MM:%g ACCEL:%u BR:%d RC:%d OV:%d\n",
GCodeProcessor::reserved_tag(GCodeProcessor::ETags::PA_Change).c_str(),
m_writer.filament()->id(),
pa_config_index,
_mm3_per_mm,
acceleration_i,
((path.role() == erBridgeInfill) ||(path.role() == erOverhangPerimeter)),
@@ -9126,7 +9114,7 @@ std::string GCode::_extrude(const ExtrusionPath &path, std::string description,
}
sprintf(buf, ";%sT%u MM3MM:%g ACCEL:%u BR:%d RC:%d OV:%d\n",
GCodeProcessor::reserved_tag(GCodeProcessor::ETags::PA_Change).c_str(),
m_writer.filament()->id(),
pa_config_index,
_mm3_per_mm,
acceleration_i,
((path.role() == erBridgeInfill) ||(path.role() == erOverhangPerimeter)),
@@ -9679,33 +9667,29 @@ void GCode::update_placeholder_parser_with_variant_params()
if (num_filaments == 0)
return;
// Helpers: remap config arrays from variant index space to filament_id index space.
// Helper: remap config arrays from variant index space to filament_id index space.
// After remapping, gcode templates can use param[filament_id] directly.
auto remap_floats_by_filament = [&](const auto &src) {
std::vector<double> dst(num_filaments);
for (size_t i = 0; i < num_filaments; ++i)
dst[i] = src.get_at(get_filament_config_index(i));
return dst;
};
auto remap_ints_by_filament = [&](const auto &src) {
std::vector<int> dst(num_filaments);
auto remap_by_filament = [&](const auto &src) {
std::decay_t<decltype(src.values)> dst(num_filaments);
for (size_t i = 0; i < num_filaments; ++i)
dst[i] = src.get_at(get_filament_config_index(i));
return dst;
};
// --- filament_options_with_variant: gcode indexes by filament_id ---
this->placeholder_parser().set("filament_max_volumetric_speed", new ConfigOptionFloats(remap_floats_by_filament(m_config.filament_max_volumetric_speed)));
this->placeholder_parser().set("filament_pre_cooling_temperature", new ConfigOptionInts(remap_ints_by_filament(m_config.filament_pre_cooling_temperature)));
this->placeholder_parser().set("filament_pre_cooling_temperature_nc", new ConfigOptionInts(remap_ints_by_filament(m_config.filament_pre_cooling_temperature_nc)));
this->placeholder_parser().set("filament_cooling_before_tower", new ConfigOptionFloats(remap_floats_by_filament(m_config.filament_cooling_before_tower)));
this->placeholder_parser().set("nozzle_temperature_initial_layer", new ConfigOptionInts(remap_ints_by_filament(m_config.nozzle_temperature_initial_layer)));
this->placeholder_parser().set("nozzle_temperature", new ConfigOptionInts(remap_ints_by_filament(m_config.nozzle_temperature)));
this->placeholder_parser().set("filament_max_volumetric_speed", new ConfigOptionFloats(remap_by_filament(m_config.filament_max_volumetric_speed)));
this->placeholder_parser().set("filament_pre_cooling_temperature", new ConfigOptionInts(remap_by_filament(m_config.filament_pre_cooling_temperature)));
this->placeholder_parser().set("filament_pre_cooling_temperature_nc", new ConfigOptionInts(remap_by_filament(m_config.filament_pre_cooling_temperature_nc)));
this->placeholder_parser().set("filament_cooling_before_tower", new ConfigOptionFloats(remap_by_filament(m_config.filament_cooling_before_tower)));
this->placeholder_parser().set("nozzle_temperature_initial_layer", new ConfigOptionInts(remap_by_filament(m_config.nozzle_temperature_initial_layer)));
this->placeholder_parser().set("nozzle_temperature", new ConfigOptionInts(remap_by_filament(m_config.nozzle_temperature)));
// first_layer_temperature is a legacy alias of nozzle_temperature_initial_layer
this->placeholder_parser().set("first_layer_temperature", new ConfigOptionInts(remap_ints_by_filament(m_config.nozzle_temperature_initial_layer)));
this->placeholder_parser().set("first_layer_temperature", new ConfigOptionInts(remap_by_filament(m_config.nozzle_temperature_initial_layer)));
this->placeholder_parser().set("pressure_advance", new ConfigOptionFloats(remap_by_filament(m_config.pressure_advance)));
this->placeholder_parser().set("enable_pressure_advance", new ConfigOptionBools(remap_by_filament(m_config.enable_pressure_advance)));
// --- printer_options_with_variant_1: in m_config these are already merged as filament-indexed ---
this->placeholder_parser().set("retraction_distances_when_cut", new ConfigOptionFloats(remap_floats_by_filament(m_config.retraction_distances_when_cut)));
this->placeholder_parser().set("retraction_distances_when_cut", new ConfigOptionFloats(remap_by_filament(m_config.retraction_distances_when_cut)));
// hotend_cooling_rate / hotend_heating_rate: gcode uses [filament_map[x]-1] (extruder_id), no remap needed
// --- filament_map: per-layer dynamic, sync from m_config to placeholder_parser ---
@@ -9715,10 +9699,10 @@ void GCode::update_placeholder_parser_with_variant_params()
{
// Fast purge mode uses filament_flush_temp_fast; Default is inert.
bool use_fast_flush = m_config.prime_volume_mode == PrimeVolumeMode::pvmFast;
auto flush_v_speed = remap_floats_by_filament(m_config.filament_flush_volumetric_speed);
auto filament_max_v = remap_floats_by_filament(m_config.filament_max_volumetric_speed);
auto flush_temps = remap_ints_by_filament(use_fast_flush ? m_config.filament_flush_temp_fast
: m_config.filament_flush_temp);
auto flush_v_speed = remap_by_filament(m_config.filament_flush_volumetric_speed);
auto filament_max_v = remap_by_filament(m_config.filament_max_volumetric_speed);
auto flush_temps = remap_by_filament(use_fast_flush ? m_config.filament_flush_temp_fast
: m_config.filament_flush_temp);
for (size_t i = 0; i < num_filaments; ++i) {
if (flush_v_speed[i] == 0)
flush_v_speed[i] = filament_max_v[i];
@@ -9770,17 +9754,11 @@ std::string GCode::set_extruder(unsigned int new_filament_id, double print_z, bo
gcode += this->placeholder_parser_process("filament_start_gcode", filament_start_gcode, new_filament_id, &config);
check_add_eol(gcode);
}
if (m_config.enable_pressure_advance.get_at(new_filament_id)) {
// In IMEX parallel modes each carriage needs an explicit tool address.
// In primary mode (single active tool) regular tool changes handle PA
// so no qualifier is needed — same as a non-IMEX printer.
const int pa_tool = imex_pem_tool_for((int)new_filament_id, m_imex_parallel_mode, m_config.physical_extruder_map);
gcode += m_writer.set_pressure_advance(m_config.pressure_advance.get_at(new_filament_id), pa_tool);
// Orca: Adaptive PA
// Reset Adaptive PA processor last PA value
m_pa_processor->resetPreviousPA(m_config.pressure_advance.get_at(new_filament_id));
}
// In IMEX parallel modes each carriage needs an explicit tool address. In primary
// mode, and on a non-IMEX printer, imex_pem_tool_for() returns -1 and the qualifier
// is omitted, because regular tool changes already handle PA there.
gcode += set_filament_pressure_advance(new_filament_id,
imex_pem_tool_for((int) new_filament_id, m_imex_parallel_mode, m_config.physical_extruder_map));
// Same suppression as the long multi-extruder path: at print-start in IMEX
// parallel modes the user's mode_gcode + machine_start_gcode is responsible
@@ -10205,19 +10183,28 @@ std::string GCode::set_extruder(unsigned int new_filament_id, double print_z, bo
if (m_ooze_prevention.enable && !defer_temp_wait)
gcode += m_ooze_prevention.post_toolchange(*this);
if (m_config.enable_pressure_advance.get_at(new_filament_id)) {
const int pa_tool = imex_pem_tool_for((int)new_filament_id, m_imex_parallel_mode, m_config.physical_extruder_map);
gcode += m_writer.set_pressure_advance(m_config.pressure_advance.get_at(new_filament_id), pa_tool);
// Orca: Adaptive PA
// Reset Adaptive PA processor last PA value
m_pa_processor->resetPreviousPA(m_config.pressure_advance.get_at(new_filament_id));
}
// See the note on the single-extruder path above: -1 outside IMEX parallel modes.
gcode += set_filament_pressure_advance(new_filament_id,
imex_pem_tool_for((int) new_filament_id, m_imex_parallel_mode, m_config.physical_extruder_map));
//Orca: tool changer or IDEX's firmware may change Z position, so we set it to unknown/undefined
m_last_pos_defined = false;
return gcode;
}
std::string GCode::set_filament_pressure_advance(unsigned int filament_id, int tool)
{
const size_t fi = get_filament_config_index(filament_id);
if (!m_config.enable_pressure_advance.get_at(fi))
return {};
// Orca: Adaptive PA
// Reset Adaptive PA processor last PA value
m_pa_processor->resetPreviousPA(m_config.pressure_advance.get_at(fi));
// tool >= 0 addresses one IMEX carriage explicitly; -1 omits the qualifier, so every
// caller that does not pass one emits exactly what it did before.
return m_writer.set_pressure_advance(m_config.pressure_advance.get_at(fi), tool);
}
inline std::string polygon_to_string(const Polygon &polygon, Print *print, bool is_print_space = false) {
std::ostringstream gcode;
gcode << "[";