Add support for direct Jaqal ingestion with `JaqalService()` by bharat-thotakura · Pull Request #1364 · Infleqtion/client-superstaq

bharat-thotakura · 2026-03-31T21:52:46Z

No description provided.

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

Copilot

Pull request overview

This PR adds first-class support for compiling Jaqal program strings directly via a new Service.jaqal_compile() API, while also refactoring shared QSCOUT option validation and consolidating duplicated CompilerOutput logic into general_superstaq.

Changes:

Add Service.jaqal_compile() to compile Jaqal strings for QSCOUT targets, including ECA support.
Centralize QSCOUT option validation in general_superstaq.validation.get_validated_qscout_options() and reuse it from Cirq/Qiskit QSCOUT compile entrypoints.
Introduce general_superstaq.compiler_output.BaseCompilerOutput and migrate Cirq/Qiskit CompilerOutput to inherit from it.

Reviewed changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated 7 comments.

Show a summary per file

File	Description
qiskit-superstaq/qiskit_superstaq/superstaq_backend.py	Refactors QSCOUT option-building to use shared validation helper.
qiskit-superstaq/qiskit_superstaq/compiler_output.py	Switches `CompilerOutput` to inherit from shared base; updates Jaqal subcircuit aggregation call site.
general-superstaq/general_superstaq/validation.py	Adds Jaqal qubit inference and a shared QSCOUT options validator.
general-superstaq/general_superstaq/validation_test.py	Adds unit tests for Jaqal qubit inference.
general-superstaq/general_superstaq/service.py	Adds `Service.jaqal_compile()` for direct Jaqal ingestion/compilation.
general-superstaq/general_superstaq/service_test.py	Adds tests covering `jaqal_compile()` behaviors (single, list, ECA, invalid inputs).
general-superstaq/general_superstaq/compiler_output.py	New shared `BaseCompilerOutput` and Jaqal-specific `read_json_jaqal()`.
general-superstaq/general_superstaq/compiler_output_test.py	Adds tests for new compiler output behavior (repr/eq basics).
general-superstaq/general_superstaq/init.py	Exports `BaseCompilerOutput` and `compiler_output` module.
cirq-superstaq/cirq_superstaq/service.py	Refactors QSCOUT option-building to use shared validation helper.
cirq-superstaq/cirq_superstaq/compiler_output.py	Switches `CompilerOutput` to inherit from shared base; updates Jaqal subcircuit aggregation call site.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

bharat-thotakura · 2026-04-08T16:36:23Z

Discussion from dev meeting: instead of jaqal_compile() which could propagate to css and qss, consider isolating with some JaqalService() whose sole job is dealing with Jaqal instead of overloading gss.Service(). For future outlook, this could also mean QASMService() and QIRService()

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

richrines1 · 2026-04-16T16:05:56Z

+                `jaqal_programs`.
+        """
+        target = gss.validation.validate_target(target)
+        circuits_is_list = not isinstance(jaqal_programs, str)


if this is a string could it still contain subcircuits? if so maybe we just make this

Suggested change

circuits_is_list = not isinstance(jaqal_programs, str)

circuits_is_list = True

to avoid randomly changing return types between one and two+ subcircuits

Currently, the server will always return a list, so if there were subcircuits, that is handled via the check in read_json_jaqal:

client-superstaq/general-superstaq/general_superstaq/compiler_output.py

Line 223 in 2fccf75

if circuits_is_list or len(compiled_circuits) != 1 or num_eca_circuits:

If circuits_is_list is always True, we wouldn't be able to use the singular .circuit attribute right?

ah, missed the check in read_json_jaqal, and also forgot we provide .jaqal_program(s) regardless of circuits_is_list (though maybe that raises a question about whether the compiler output from jaqal service should just use .circuit(s) instead of .jaqal_program(s)?)

we wouldn't be able to use the singular .circuit attribute right?

my thinking was that this would make sense in this case, because if i pass a single jaqal string there's no way to tell if i intended it to be a standalone circuit or a program with a single subcircuit (and imo it's not great to change switch just given the number of circuits, because it makes the result type inconsistent in any sort of scripting). but feel less strongly due to the existence of .jaqal_program(s) (and the coming deprecation of compiler outputs anyway)

I initially went with circuits_is_list = True, so I'm not opposed to it since the attribute change isn't elegant (I think I didn't want to index by 0 for single-str inputs). However, in light of the review, I've just made it always a list with 77ca880. Let me know if that works better

richrines1 · 2026-04-16T16:10:59Z

+Q = TypeVar("Q")
+
+
+class BaseCompilerOutput(Generic[C, Q]):  # noqa: PLW1641


love the use of generics here - wondering if we could do something similar for BaseService (obviously not for this pr though)

Yeah probably, but feels like we would end up implementing the specific overrides anyway (e.g., because of different serializations etc), so would have to see how much more is consolidated and not duplicated. Regardless, will leave for another PR when the SuperstaqClientV3 becomes the norm

ah good point, that's probably true

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

richrines1

lgtm! (assuming tests will pass after #1380)

bharat-thotakura · 2026-04-17T23:41:15Z

(assuming tests will pass after #1380)

Yup, they should (except for the unrequired notebook checks)

bharat-thotakura added 8 commits March 31, 2026 16:50

factor out CompilerOutput into gss

54fec83

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

add compiler_output.py to gss

6577c55

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

add jaqal_compile()

c1ae787

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

add license header

dc120d7

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

handle some subcircuits

0b478a6

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

remove unecessary check + refactor

f30dca8

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

add tests for coverage

d3dc09c

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

Merge branch 'main' into bt/support-jaqal-input

b45e4a8

bharat-thotakura marked this pull request as ready for review April 6, 2026 22:55

bharat-thotakura marked this pull request as draft April 6, 2026 22:58

bharat-thotakura added 4 commits April 6, 2026 18:03

type check fix

27756d7

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

refactor common qscout options

e6f7a73

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

add gss compiler output tests

7d9b31a

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

unit tests for coverage

d6982cb

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

bharat-thotakura marked this pull request as ready for review April 7, 2026 16:16

nits

e5c1c43

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

bharat-thotakura requested a review from Copilot April 7, 2026 16:33

Copilot started reviewing on behalf of bharat-thotakura April 7, 2026 16:34 View session

Copilot AI reviewed Apr 7, 2026

View reviewed changes

bharat-thotakura marked this pull request as draft April 7, 2026 19:41

bharat-thotakura and others added 7 commits April 7, 2026 15:40

account for ECA properly

8efc41d

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

allow integer validation to print name of variable

9003d5a

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

update affected tests

04e71d2

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

Correct class name in __repr__

bef3933

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

jaqal_compile() docstring nit

805f0e7

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

update

030f99e

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

revert eq check

1237576

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

bharat-thotakura marked this pull request as ready for review April 7, 2026 21:57

Merge branch 'main' into bt/support-jaqal-input

f0bd41d

bharat-thotakura marked this pull request as draft April 8, 2026 21:41

bharat-thotakura and others added 4 commits April 8, 2026 17:16

Merge branch 'main' into bt/support-jaqal-input

e2659f6

refactor to use dedicated JaqalService

3487b2b

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

docstring nits

e4fd762

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

nits

777422f

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

bharat-thotakura marked this pull request as ready for review April 10, 2026 20:28

Merge branch 'main' into bt/support-jaqal-input

2fccf75

richrines1 reviewed Apr 16, 2026

View reviewed changes

bharat-thotakura added 2 commits April 16, 2026 11:41

change compile to qscout compile

4f8991c

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

treat circuit_is_list as always True for Jaqal input

77ca880

Signed-off-by: Bharath <bharath.thotakura@infleqtion.com>

bharat-thotakura requested a review from richrines1 April 16, 2026 21:16

bharat-thotakura changed the title ~~Add support for direct Jaqal ingestion with jaqal_compile()~~ Add support for direct Jaqal ingestion with JaqalService() Apr 17, 2026

richrines1 approved these changes Apr 17, 2026

View reviewed changes

Merge branch 'main' into bt/support-jaqal-input

7a88422

bharat-thotakura merged commit de2c7d4 into main Apr 20, 2026
18 of 21 checks passed

bharat-thotakura deleted the bt/support-jaqal-input branch April 20, 2026 20:07

	circuits_is_list = not isinstance(jaqal_programs, str)
	circuits_is_list = True

		Q = TypeVar("Q")


		class BaseCompilerOutput(Generic[C, Q]): # noqa: PLW1641

Uh oh!

Conversation

bharat-thotakura commented Mar 31, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bharat-thotakura commented Apr 8, 2026

Uh oh!

richrines1 Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

bharat-thotakura Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

richrines1 Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

bharat-thotakura Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

richrines1 Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

bharat-thotakura Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

richrines1 Apr 17, 2026

Choose a reason for hiding this comment

Uh oh!

richrines1 left a comment

Choose a reason for hiding this comment

Uh oh!

bharat-thotakura commented Apr 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

bharat-thotakura commented Apr 17, 2026 •

edited

Loading