Store schedule and pulse history details from output for post processing by anu1217 · Pull Request #240 · svalinn/ALARA

anu1217 · 2026-02-03T21:35:53Z

This script writes the schedule and pulsing history related output to dictionaries so that they can be accessed and manipulated in other scripts. The schedule details are written in a nested dictionary to accommodate sub-schedules.

fixes #220

add more logic more logic store output combine multiple commits into one

gonuke

There is a repeated suggestion here to use a while loop instead of a for loop when looking through the lines. This is a typical design pattern for a parser.

A "design pattern" is typical and usually well-tested algorithm pattern for accomplishing a given task. Design patterns transcend languages and when you use them, others can quickly grasp what the intent of the code is and not worry about understanding the high level algorithm, but quickly focus on the details.

A "parser" is a typical pattern for moving through a stream of tokens (lines in your case) with two primary things that can happen at each line:

change the "state"
take action depending on the state

Zero, one, or both of these things can happen at each new token.

The most common action is to process the value of a token into some other data structure.

gonuke · 2026-02-04T11:55:48Z

+
+def read_out(output_path):
+    with open(output_path, "r") as output_file:
+        lines = output_file.readlines()


Just learned that this might make life easier by stripping all the lines when you read:

Suggested change

lines = output_file.readlines()

lines = output_file.read().striplines()

Hmm...output_file.read() returns a string that has no method called striplines(). I also can't seem to find any documentation on this method?

My bad... it's splitlines https://www.w3schools.com/python/ref_string_splitlines.asp

gonuke · 2026-02-04T11:59:20Z

+    matches = []
+    for key, value in top_dict.items():
+
+        if key.startswith(search_str):


At this point, haven't you parsed things such that this should be an equality test rather than a startswith() test?

This is because of the schedule ... lines. Because the name of the schedule is not known beforehand, it is not possible to find an exact match. At first, I set each key to be just "schedule" without the name, but in that case make_nested_dict() will only store the last schedule line.

gonuke · 2026-02-04T12:09:00Z

+                     }
+    search_fields = ["schedule", "pulse_entry:"]
+    for search_field in search_fields:
+        matches = search_for_match(sch_dict, search_field)


search_for_match seems to be designed to return a list, but this method seems to assume that only value will be found. Is that right?

Initially I set this up such that the list was essentially a dummy list with only one entry, but this was not correctly capturing all of the entries. The list now contains multiple entries, which are values corresponding to each of the search field keys. The loop also iterates over each of the entries in the list.

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

change order of multiplication fix indentation modify logic for counter further modify counter correctly increment counter when pulse_entry is on the same indentation level as schedule combine similar changes into one commit

anu1217 · 2026-02-05T08:16:37Z

I realized that the counter I had originally wasn't correctly keeping track of pulse entries when they were on the same indentation level as the schedule lines. The new counter should be able to account for those cases.

gonuke

I think make_nested_dict is more complicated than necessary- see my comments

gonuke · 2026-02-05T11:47:15Z

+    unit_multipliers = make_unit_multipliers()
+    while not lines[line_idx].startswith("pulse_history:"): # next section of output
+        child_level = lines[line_idx].count('\t')
+        newline_name = lines[line_idx].strip().split()


This variable name doesn't really seem to express what this data is...

gonuke · 2026-02-05T12:01:07Z

+            last_upper_indent_level[child_level][f"pulse_entry num_{counter}_in_sched"] = make_pe_sub_dict(newline_name,
+                                                                                                           unit_multipliers)  
+            last_upper_indent_level[child_level + 1] = last_upper_indent_level[child_level][f"pulse_entry num_{counter}_in_sched"]
+            counter = counter_dict[child_level] = counter + 1


I don't think you need to set counter here... you never use it before it gets reassigned

gonuke · 2026-02-05T12:01:35Z

+            last_upper_indent_level[child_level + 1] = last_upper_indent_level[child_level][newline_name[1].strip("':")]           
+            line_idx += 1
+            new_child_level = lines[line_idx].count('\t')
+            counter = counter_dict[new_child_level] = 1


You don't need counter here... you never use it before it gets reassigned

gonuke · 2026-02-05T12:04:06Z

+            last_upper_indent_level[child_level + 1] = last_upper_indent_level[child_level][f'schedule {newline_name[1]}']
+            line_idx += 1 
+            new_child_level = lines[line_idx].count('\t')
+            counter = counter_dict[new_child_level] = 1


Is this based on the assumption that this will be the first time we encounter this child_level such that we are initializing this counter for the first time? Or are we also resetting this counter under some circumstances?

I think it's the latter, in which case I would do this elsewhere. Also - why not initialize with 0?

gonuke · 2026-02-05T12:05:59Z

+    counter_dict = {}
+    unit_multipliers = make_unit_multipliers()
+    while not lines[line_idx].startswith("pulse_history:"): # next section of output
+        child_level = lines[line_idx].count('\t')


If my understanding about the counters is correct, you can initialize new counters here by seeing whether a counter already exists at this child_level and initializing one if it doesn't, and then avoid doing it in two places below.

anu1217 · 2026-02-05T22:47:33Z

To explain the idea behind the counters further...I'm initializing it to 1 instead of 0 so that we can track the order of the pulse entries belonging to a schedule in a way that's perhaps more human readable.

There are a few different scenarios where this counter needs to be set/reset.

a) The else condition initializes the counter to 1 at child_level 1 to account for the case where a pulse entry is part of the top schedule and not any sub-schedule.
b) Whenever a schedule line is found, the counter is reset at the line following it, since a pulse_entry must follow the schedule line.
c) The counter is reset at the line after schedule, and not at schedule itself, to avoid changing the pulse entry count at the indentation level of schedule. As an example, in:

top ...
  pulse_entry
  sch ..
      pulse_entry
  pulse_entry

resetting the counter at the 3rd line would mean that the last pulse entry is incorrectly counted as the 1st pulse in the top schedule.

This did take some trial and error to figure out, and would probably be a good discussion topic for hacker-within.

This reverts commit 7a40d56.

This reverts commit 95a4f37.

New design structure for nested dictionary in svalinn#240 (svalinn/ALARA)

gonuke

Here is a coordinated set of changes to streamline the algorithm. It is designed to always initialize things correctly rather than have to check later to do that.

always put a "children" entry in the dictionary when we make it
initialize ancestors rather than wait to find "top_schedule"

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

gonuke

This seems OK now, but still probably a little unusual in the way the parsing is put together.

anu1217 added 3 commits January 30, 2026 16:20

add missing '

8350071

change output formatting

c9ed99f

skeleton script for post-processor

f7b922e

add more logic more logic store output combine multiple commits into one

anu1217 changed the title ~~Store output details for post processing~~ Store schedule and pulse history details from output for post processing Feb 3, 2026

anu1217 and others added 3 commits February 3, 2026 15:38

remove extra line

d07763b

fix order of function execution

301c46f

store all entries correctly

3f52086

gonuke requested changes Feb 4, 2026

View reviewed changes

gonuke mentioned this pull request Feb 4, 2026

Create tool to plot simple irradiation histories #238

Open

anu1217 and others added 16 commits February 4, 2026 11:04

Use line.count() to find # of tabs

c18ff44

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

fix typo

c78548c

actually store all pulse entries

c25e91f

fix typo

52a2083

Use while loop to read lines

d517afe

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

change key entries

99875b5

make helper functions for cases

f8b583d

Use while loop for cases

95e5e7a

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

use while loop to parse lines

ea7c96c

use newline_name throughout

d3df115

Update tools/sched_post_processor.py

79cd56d

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

change some variable names

28865bf

convert units in make_nested_dict()

98639fd

remove unused function

a26c428

change order of multiplication

795bc22

strip and split lines before checking cases

76d8f9d

change order of multiplication fix indentation modify logic for counter further modify counter correctly increment counter when pulse_entry is on the same indentation level as schedule combine similar changes into one commit

anu1217 force-pushed the post_proc_sched branch from 8694e69 to 76d8f9d Compare February 5, 2026 08:13

gonuke requested changes Feb 5, 2026

View reviewed changes

anu1217 added 2 commits February 5, 2026 15:31

remove unit multipliers from function

5064d4f

black format document

d407b54

anu1217 added 3 commits February 5, 2026 15:49

yapf format document

b23c6b9

reassign keys in counter_dict directly

51446fc

change name of last_upper_indent_level

da33c51

anu1217 and others added 14 commits February 6, 2026 14:59

write sch dict to json file

95a4f37

reformat document

7a40d56

Revert "reformat document"

0f94b6a

This reverts commit 7a40d56.

Revert "write sch dict to json file"

3ab5901

This reverts commit 95a4f37.

move unit multipliers to make_nested_dict()

40d0cc3

reformat document

41363de

add extra colon

a0105c1

distinguish top schedule from sub schedules

77b1f0a

change output formatting

0b1a243

fix merge conflict

1350cbc

change condition for finding lines

27aea09

alternative parsing scheme

6cf0a88

trying some new ideas

3617b23

initialize current_sched[children and account for top schedule

5e6a004

anu1217 mentioned this pull request Feb 12, 2026

New design structure for nested dictionary in #240 (svalinn/ALARA) gonuke/ALARA#1

Merged

Merge pull request #1 from anu1217/mod_post_proc_sched

3bad66f

New design structure for nested dictionary in svalinn#240 (svalinn/ALARA)

gonuke requested changes Feb 17, 2026

View reviewed changes

Comment thread tools/sched_post_processor.py Resolved

Comment thread tools/sched_post_processor.py Resolved

Comment thread tools/sched_post_processor.py Resolved

Comment thread tools/sched_post_processor.py Resolved

Comment thread tools/sched_post_processor.py Resolved

anu1217 and others added 5 commits February 17, 2026 10:46

Make "children" key inside sched_tree

cac4f75

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

Initialize "children" key inside sched line dict

c1363a8

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

Remove unused line

5bd4591

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

Append current_sched to ancestors during initialization

49ef327

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

Remove unused line

506f332

Co-authored-by: Paul Wilson <paul.wilson@wisc.edu>

gonuke approved these changes Feb 17, 2026

View reviewed changes

gonuke merged commit 5e9e872 into svalinn:main Feb 17, 2026
1 check passed

anu1217 deleted the post_proc_sched branch February 24, 2026 19:52

	lines = output_file.readlines()
	lines = output_file.read().striplines()

Conversation

anu1217 commented Feb 3, 2026

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

anu1217 commented Feb 5, 2026

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

anu1217 commented Feb 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

anu1217 commented Feb 5, 2026 •

edited

Loading