Skip to content

Commit

Permalink
update capability, etc
Browse files Browse the repository at this point in the history
  • Loading branch information
Fry authored and Fry committed Mar 25, 2017
1 parent 097dcef commit d4c6c80
Show file tree
Hide file tree
Showing 16 changed files with 475 additions and 128 deletions.
4 changes: 2 additions & 2 deletions dextersim.js
View file
Original file line number Diff line number Diff line change
Expand Up @@ -115,13 +115,13 @@ DexterSim = class DexterSim{
DexterSim.fill_in_robot_status_joint_angles(robot_status)
*/
break;
case "B": //move_to_relative xyz
// case "B": //move_to_relative xyz
/*if (!isNaN(instruction_array[2])) robot_status[Dexter.ds_j5_x_index] = robot_status[Dexter.ds_j5_x_index] + instruction_array[2]
if (!isNaN(instruction_array[3])) robot_status[Dexter.ds_j5_y_index] = robot_status[Dexter.ds_j5_y_index] + instruction_array[3]
if (!isNaN(instruction_array[4])) robot_status[Dexter.ds_j5_z_index] = robot_status[Dexter.ds_j5_z_index] + instruction_array[4]
if (!isNaN(instruction_array[5])) robot_status[Dexter.ds_j4_angle_index] = instruction_array[5] //not relative
DexterSim.fill_in_robot_status_joint_angles(robot_status)*/
break;
// break;
case "e": //cause an error. Used for testing only
robot_status[Dexter.ERROR_CODE] = instruction_array[2]
break;
Expand Down
221 changes: 191 additions & 30 deletions doc/ref_man.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -87,6 +87,41 @@

<details style="margin-left:20px;"><summary>Sound</summary>

<details id="beep_doc_id" style="margin-left:20px;"><summary>beep</summary>
Plays an electronic sound.<br/>
<i>Parameters</i>:<br/>
<b>duration</b> A number milliseconds. Default: 500<br/>
<b>frequency</b> A number in Hertz. Default: 400<br/>
<b>volume</b> A number between zero and 1. Default: 1<br/>
<b>waveform</b> The timbre or waverform of the sound. Default: "triangle".
Other options are: "sine", "square", "sawtooth".<br/>
<b>callback</b> A function of no arguments that is called when the beep finishes playing.
Default: null, meaning do nothing.<br/>
<i>Example</i>: <pre>
beep({duration: 500,
frequency: 440,
volume: 1,
waveform: "triangle",
callback: function(){beep({frequency:493.88})}
})
</pre>
</details>

<details id="beeps_doc_id" style="margin-left:20px;"><summary>beeps</summary>
Beeps the indicated number of times with the default <code>beep</code> settings.<br/>
<i>Parameters</i>:<br/>
<b>times</b> An integer indicated the number of times to beep. Default: 1.<br/>
<b>callback</b> A function of no arguments that is called when beeps finishes playing.
Default: null, meaning do nothing.<br/>
<i>Example</i>: <pre>
beeps(3,
function(){
speak({speak_data: "Third Floor, home robots"})
})
</pre>
</details>


<details id="recognize_speech_doc_id" style="margin-left:20px;"><summary>recognize_speech</summary>
Speech recognition has a long history in AI. After many decades of research, it <i>can</i>
be useful, but don't expect it to be as reliable as a human. DDE uses Google's speech
Expand Down Expand Up @@ -437,6 +472,14 @@
The default is 0. You can use any instruction_location here. See Robot.go_to
for documentation on all the different kinds of instruction_locations.
</li>
<li><b>ending_program_counter</b> Declare at the outset of a job that it
will not execute the indicated instruction or beyond. This can be any
valid instruction_location that does not include a job. It is evaluated
at the start of each instruction, so if labels move as a result of
dynamically added instructions, that's OK.
The default is "end". See <code>Robot.go_to</code>
for documentation on all the different kinds of instruction_locations.
</li>
<li><b>initial_instruction</b> Add an initial instruction to the do_list at
the location of program_counter. Thus this instruction will be executed
first when the job is run. Default null (no instruction added).
Expand Down Expand Up @@ -502,6 +545,72 @@
<details style="margin-left:20px;"><summary>Robot Instructions</summary>
These instructions can be used for any robot, as can the instructions for
the Human robot.
<details style="margin-left:20px;"><summary>Any Function Def</summary>
Any function definition can be put on the do_list as an instruction.
When it is run, the function is called with <code>this</code>
being the instance of Job that the function def is in. Since you can put any
JavaScript inside a function definition, you can have an instruction
execute any JS whatsoever.
<p/>
Often you'll make functions that have no returned value.
If so, after the function is called, then the next instruction
on the do_list will be run. But if the function returns
another instruction, or an array of instructions, then those
will be inserted into the do_list immediately below the
function def, and they will be run next.
<p/>
<i>The below functionality is advanced use of DDE.</i><br/>
The function
body might have an <code>if</code> in it. Under certain
conditions the function will return one instruction,
and it other cases, it will return another (or <code>null</code>, meaning
just do the next instruction on the do_list, or an array of
instructions to do them all).
<p/>
If a function returns the special string
<code>"dont_call_set_up_next_do"</code>,
then that will cause nothing to happen in the job. It won't advance.
This is not a good thing to do by itself, but you might
want to conditionally call
<code>this.set_up_next_do()</code>. This tells DDE to
continue processing this job's instructions.
The argument to this function is an integer saying how much
to increment the program counter. It defaults to 1, meaning
do the next instruction. Another useful value is 0,
which will mean that after a short pause, run this instruction
again. Then in some other running of the instruction, a different
branch of an <code>if</code> may be taken that will call
<code>this.set_up_next_do(1)</code> causing the job to proceed.
<i>Example</i>:<pre>
function pause_a_bit(){
if (new Date().getSeconds() < 30){
this.set_up_next_do(0)
}
else {
this.set_up_next_do(1)
}
return "dont_call_set_up_next_do"
}
</pre>
The above example could better be implemented using
<code>Robot.wait_until</code>, but there are
cases where the increased flexibility of the above is convenient.
<p/>
You can put both named and anonymous functions on a do_list.
A named instruction allows you to add a bit of "documentation" to
the function. You can also use the name of a function like a
label and "go to"
that name using a <code>Robot.go_to</code> instruction,
or use the function's name in
any uses of an instruction_location.
<p/>
If you name your function definition,
you can define it <i>outside</i> of the do_list, then just
have its name in the do_list source code. This lets you put
the same function in multiple locations on the do_list and
only have to define it once.
</details>


<details id="Robot.error_doc_id" style="margin-left:20px;"><summary>error</summary>
Causes the job to error, immediately stopping it.
Expand Down Expand Up @@ -565,6 +674,7 @@
instruction, etc.<br/>
<b>the name of a label instruction</b> The program counter is set to that label's id.<br/>
<b>the name of a sync_point instruction</b> The program counter is set to that instruction's id.<br/>
<b>the name of a function def instruction</b> The program counter is set to that instruction's id.<br/>
<b>the type of instruction</b> Examples: <code>"go_to", "label", "sync_point"</code> etc.
For example, if we have "label", then we're not going to a specific label, just
the next instruction of type "label" found.<br/>
Expand Down Expand Up @@ -731,6 +841,25 @@
See Job Example 9 for an extended example.
</details> <!-- end grab_robot_status -->

<details id="Robot.if_any_errors_doc_id" style="margin-left:20px;"><summary>if_any_error</summary>
<code>if_any_error</code> checks to see if any of the supplied jobs have errored.
If so, it inserts the supplied instruction into the do_list on the job
that this instruction is on. In either case, <code>if_any_error</code> continues
the job that its on.
<i>Parameters</i><br/>
<b>job_names</b> An array of strings of job names. If a named job doesn't exist, error.<br/>
<b>instruction_if_error</b> An instruction that will be inserted into the do_list
immediately after this instruction IF one of the named jobs has errored.
The default for <code>instruction_if_error</code> is a function that
inserts an instance of Robot.error into the do_list of this job, causing it to error.
The "reason" for that error instruction is explicit about the cause of the error.
<p/>
Generally a job should be stopped quickly if there is an error.
This instruction makes it easy to stop jobs that are concurrently cooperating
with the current job. The instruction_if_error can be a function (called with the
current job for "this") so that some arbitrary compensation can be made for the error.
</details>

<details id="Robot.label_doc_id" style="margin-left:20px;"><summary>label</summary>
This instruction is a no-op. However, it provides a target location
for <code>Robot.go_to</code>, initializing the program_counter of a job, and
Expand All @@ -757,32 +886,48 @@
its job name and instruction ID right before printing <code>val</code>.
</details>

<details id="Robot.if_any_errors_doc_id" style="margin-left:20px;"><summary>if_any_error</summary>
<code>if_any_error</code> checks to see if any of the supplied jobs have errored.
If so, it inserts the supplied instruction into the do_list on the job
that this instruction is on. In either case, <code>if_any_error</code> continues
the job that its on.
<i>Parameters</i><br/>
<b>job_names</b> An array of strings of job names. If a named job doesn't exist, error.<br/>
<b>instruction_if_error</b> An instruction that will be inserted into the do_list
immediately after this instruction IF one of the named jobs has errored.
The default for <code>instruction_if_error</code> is a function that
inserts an instance of Robot.error into the do_list of this job, causing it to error.
The "reason" for that error instruction is explicit about the cause of the error.
<p/>
Generally a job should be stopped quickly if there is an error.
This instruction makes it easy to stop jobs that are concurrently cooperating
with the current job. The instruction_if_error can be a function (called with the
current job for "this") so that some arbitrary compensation can be made for the error.
</details>


<details id="Robot.stop_doc_id" style="margin-left:20px;"><summary>stop</summary>
Causes the job to stop. This is an expected stop, as constrasted with
<b>Robot.error</b>.
The one argument is a string of a reason for stopping which
will aid in debugging. Example:<pre>
Robot.stop("all done")</pre>
<details id="Robot.start_job_doc_id" style="margin-left:20px;"><summary>start_job</summary>
Starts another job.<br/>
If the status of the job to be started is:<br/>
<b>starting</b> This instruction is ignored.<br/>
<b>suspended</b> The job is unsuspended.<br/>
<b>"not_started", "completed", "errored", or "interrupted"</b>
The job is started.<br/>
<b>"running" or "waiting"</b> then what happens depends on the
value of <code>if_started</code><br/>

<i>Parameters:</i><br/>
<b>job_name</b> A string. The name of the job to start. Required.<br/>
<b>options</b> A literal object. The options to pass to the <code>start</code>
function. These are the same as the options passed to creating a job except
that <code>name</code> is ignored.<br/>
<i>Example</i>: <code>{program_counter: 2}</code> <br/>
causes the job to be started, executing its 3rd do_list item first.<br/>
<b>if_started</b> A string. This is only relevent if the job to be started
has a status of "running" or "waiting".<br/>
If if_started == <code>"ignore"</code>, then nothing changes and
the job to be started is allowed to proceed. This is the default.<br/>
If if_started == <code>"error"</code>, then the job containing
this instruction errors.<br/>
If if_started == <code>"restart"</code>, then the job to be started
is stopped and restarted.
</details>

<details id="Robot.stop_job_doc_id" style="margin-left:20px;"><summary>stop_job</summary>
Causes the indicated job to stop just before executing a particular instruction.<br/>
<i>Parameters</i>:<br/>
<b>instruction_location</b> Determines the job to stop and the instruction within that
job that it should stop before executing. If <code>instruction_location</code> does
not indicate a job, then the job that this instruction is in will be stopped.
If <code>instruction_location</code> is not given, then
the offset will be <code>"program_counter"</code>, i.e. the job will
stop before executing its next instruction. Thus the default is that
the current job will be stopped immediately. See Robot.go_to for details
on the <code>instruction_location</code> format.<br/>
<b>reason</b> A string. The reason that the job is being told to stop.<br/>
Example:<pre>
Robot.stop_job({job: "my_job", offset: "label1"})</pre>
<code>Job.my_job</code> will stop when it reaches the instruction named "label1".
</details>

<details id="Robot.suspend_doc_id" style="margin-left:20px;"><summary>suspend</summary>
Expand Down Expand Up @@ -933,6 +1078,22 @@
that doesn't contain a wait_until,
such as our first call to <code>insert_instruction</code> above.
</li>
<li><b>instruction_location</b> <i>Only applies to instruction_locations that
are arrays or literal objects.</i> Waits for the instruction location
to refer to a program_counter that is less than or equal to
the program_counter in the job referenced by the instruction_location.
This instruction waits until a job gets to a certain instruction.
This performs like a one_sided <code>Robot.sync_point</code> instruction
where the job this instruction is in will wait for the job
specified in the instruction_location, but not the other way around.
Its advantage over <code>Robot.sync_point</code> is that you don't have
to have a Robot.sync_point instruction in the job you're waiting for.
Thus you can wait for a job reaching one of its instructions
that never expected it to be waited for.<br/>
Warning: If the instruction_location specifies the job
that this <code>wait_until</code> instruction is in, or lets the instruction_location
default to that, this job could wait for itself and thus forever.</li>

</ul>
</details> <!-- end wait_until -->

Expand Down Expand Up @@ -1142,10 +1303,10 @@
J5_direction settings. Impossible settings cause an error.
</details> <!--end move_to -->

<details id="Dexter.move_to_relative_doc_id" style="margin-left:20px;"><summary>move_to_relative</summary>
<!--<details id="Dexter.move_to_relative_doc_id" style="margin-left:20px;"><summary>move_to_relative</summary>
This function is the same as <code>Dexter.move_to</code> except that the xyz values are ADDED to Dexter's
existing end effector xyz values, not taken as absolutes.
</details>
</details>-->

<details id="Dexter.move_all_joints_doc_id" style="margin-left:20px;"><summary>move_all_joints</summary>
move_all_joints is a lower level instruction than move_to, giving you the ability to dicate
Expand All @@ -1161,10 +1322,10 @@
<code>null</code> or the lack of enough numbers means keep the angle for that joint what it is at now.
</details>

<details id="Dexter.move_all_joints_relative_doc_id" style="margin-left:20px;"><summary>move_all_joints_relative</summary>
<!-- <details id="Dexter.move_all_joints_relative_doc_id" style="margin-left:20px;"><summary>move_all_joints_relative</summary>
The same as Dexter.move_all_joints except that the given angles are added to the current
angles of the robot. <code>null</code> or missing angle increments default to 0.
</details>
</details> -->

<details id="Dexter.move_home_doc_id" style="margin-left:20px;"><summary>move_home</summary>
Moves all joints to 0. J1 is facing forwards, J2, J3, J4 are straight up. J5
Expand Down
45 changes: 43 additions & 2 deletions doc/release_notes.html
View file
Original file line number Diff line number Diff line change
@@ -1,3 +1,46 @@
<details style="padding-left:20px;"><summary>v 1.0.9, Mar 25, 2017</summary>
Highlights: Updating DDE much easier. Robot.start_job and Robot.stop_job.
Improved Doc. Job init param: ending_program_counter.
<ul>
<li> When using Doc pane Find button, better help text printed in
Output pane.</li>

<li> <code>Robot.wait_until</code> can now take an instruction_location
(only arrays and lit objs format accepted) as it argument
to wait until that instruction_location is reached in another job.</li>

<li>when DDE launches it prints in the output pane its
DDE version and date. If its not the latest release,
it tells you what is and how to get it.</li>

<li> "Update..." is a new File menu item that tells you the same thing.</li>

<li> Removed DDE Dexter instructions: <code>move_all_joints_relative</code>
and <code>move_to_relative</code>. Just use the new coord sys.</li>

<li> Documented "Any Function Def" as a kind of instruction in a do_list,
under Ref Man Robot. This was an important omission: recommended reading!</li>

<li> User Guide 'Installation" now documents new github process.</li>

<li> New instruction <code>Robot.start_job</code>
let's you start other jobs flexibly from within a job.</li>

<li>New init option for a job (and the start function)
<code>ending_program_counter</code> allows you to
declare at the outset to stop a job early.
Can be any instruction_location.</li>

<li> Extended instruction: <code>Robot.stop_job</code> (renamed from Robot.stop)
Now takes an instruction_location so it can stop other jobs
AND stop them at any of their instructions.</li>

<li> <code>beep</code> and <code>beeps</code> now documented in the ref man with click help.</li>

<li> Click help on [ and ] now works.</li>
</ul>
</details>

<details style="padding-left:20px;"><summary>v 1.0.4, Mar 16, 2017</summary>
Highlights: New Electron Framework, New instruction Robot.grab_robot_status,
Serial Port extensions, file access improved, dxf improvements,
Expand Down Expand Up @@ -62,7 +105,6 @@
the url to the HDRobotic video page (under About) and the DDE video
(under "User Interface")</li>


<li> serial port improvements:
<ul><li> serial_devices() returns a list of available serial port devices.</li>
<li>serial_path_to_info_map holds an object for each connected serial port device.</li>
Expand Down Expand Up @@ -144,7 +186,6 @@
but keep it the same, and go and grab the id of the
instruction that errored and use it.</li>


<li> Instruction.Control.send_to_job no longer takes a "to_job_name"
as an explicit parameter, however,
the where_to_start param can be any instruction_location
Expand Down
2 changes: 2 additions & 0 deletions editor.js
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1321,6 +1321,8 @@ Editor.identifier_or_operator = function(){
else if (cur_char == "}") { return "}"}
else if (cur_char == "(") { return "("}
else if (cur_char == ")") { return ")"}
else if (cur_char == "[") { return "["}
else if (cur_char == "]") { return "]"}
else if ((cur_char == ".") && (pos > 0) && !is_digit(full_src[pos - 1])){ return "." } //NOT a decimal point
var bounds = Editor.bounds_of_identifier(full_src, pos)
if (bounds){
Expand Down
6 changes: 4 additions & 2 deletions eslint-lint.js
View file
Original file line number Diff line number Diff line change
Expand Up @@ -28,8 +28,10 @@
"dex": false,
"Duration": false,
"editor": false,
//"funciton*: false, doesn't work in getting rid of warning.
"get_in_ui": false,
//"function*: false, doesn't work in getting rid of warning.
"get_in_ui": false,
"get_page": false,
"get_page_async": false,
"Human": false,
"Job": false,
"out": false,
Expand Down
Loading

0 comments on commit d4c6c80

Please sign in to comment.