feat(edecimal): extend parse to decimal-point and scientific notation

[Ravenwater]

Summary

`edecimal::parse` previously matched only `[+-]*[0-9]+` via
`std::regex_match`, rejecting every form with a decimal point or an
exponent suffix. This PR routes parse through
`sw::universal::string_parse::scan_decimal_float` (the foundation from
#838) and now accepts decimal-point and scientific-notation inputs that
yield an exact integer.

Design

`scan_decimal_float` returns `int_part`, `frac_part`, and a signed
`exp10`. The value's effective decimal exponent is
`exp10 - frac_part.size()`.

• eff_exp >= 0: the value is an exact integer. Concatenate
  `int_part || frac_part`, append `eff_exp` trailing zeros, store
  the digits in edecimal's LSB-first vector. Examples:
  `"3.14e2" -> 314`, `"1.5e10" -> 15000000000`, `"3.14e+200" ->
  314 followed by 198 zeros`.
• eff_exp < 0: the value carries fractional digits that edecimal
  (an integer-only number system) cannot represent without precision
  loss. Reject. Examples: `"3.14"`, `"-0.5"`, `"1.5e-100"`,
  `"3.0"` (eff_exp = -1), `"10.50e0"` (eff_exp = -2).

The strict rejection matches the issue spec's "preserve those digits
exactly" rule.

Changes

• `include/sw/universal/number/edecimal/edecimal.hpp` -- include
  `utility/string_parse.hpp` (provides `scan_decimal_float`).
• `include/sw/universal/number/edecimal/edecimal_impl.hpp` -- rewrite
  `parse()` against `scan_decimal_float`; call `unpad()` so leading
  zeros do not survive in the digit vector; collapse negative zero to
  +0 across all accepted forms.
• `elastic/decimal/conversion/string_parse.cpp` -- extend the test
  suite to 9 groups covering the new grammar, the strict rejection,
  and the pre-existing operator>> hygiene (shipped in feat: operator>> hygiene + ereal nan/inf for decimal/elastic family (Phase E of #835) #858).

Test Results

Target	gcc build	gcc test	clang build	clang test
edec_string_parse	OK	PASS (9/9 groups)	OK	PASS (9/9 groups)
edec_api	OK	exit 0	OK	exit 0
edec_assignment	OK	exit 0	OK	exit 0
edec_comparison	OK	exit 0	OK	exit 0
edec_constexpr	OK	exit 0	OK	exit 0
edec_addition / subtraction / multiplication / division	OK	exit 0	OK	exit 0

Test plan

• Fast CI (gcc + clang CI_LITE) passes
• Promote when satisfied: `gh pr ready`

Resolves #854

Generated with Claude Code

Summary by CodeRabbit

• New Features

  • Enhanced decimal parsing: supports scientific notation, decimal points with fractional parts, and signed literals; enforces limits to prevent overly large expansions and rejects inputs that would require fractional loss.

• Tests

  • Expanded regression suite with structured per-case reporting and helpers to validate successful parses and expected rejection/fail behavior.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 9afb4d5a-a6b3-438f-8a3c-07cac4152104

📥 Commits

Reviewing files that changed from the base of the PR and between 4bd7897 and 2370c77.

📒 Files selected for processing (2)

• elastic/decimal/conversion/string_parse.cpp
• include/sw/universal/number/edecimal/edecimal_impl.hpp

🚧 Files skipped from review as they are similar to previous changes (2)

• include/sw/universal/number/edecimal/edecimal_impl.hpp
• elastic/decimal/conversion/string_parse.cpp

---

📝 Walkthrough

Walkthrough

Extended edecimal::parse() to accept decimal-point and scientific-notation literals using scan_decimal_float, reject inputs with negative effective exponent, cap expansion size, normalize digits, and expand regression tests and operator>> edge-case checks.

Changes

Extended edecimal string parsing for decimal and scientific notation

Layer / File(s)	Summary
Parse implementation with decimal and scientific notation support include/sw/universal/number/edecimal/edecimal_impl.hpp	parse() algorithm rewritten to use scan_decimal_float for tokenization, compute effective exponent (base-10 exponent minus fractional-digit count), reject negative effective exponent cases to avoid fractional loss, enforce a maximum expanded digit count, rebuild internal digit vector with significand plus exponent-derived trailing zeros, reverse to [0]==10^0, normalize via unpad(), and force zero to be positive.
Header dependency and wiring include/sw/universal/number/edecimal/edecimal.hpp	Adds #include <universal/utility/string_parse.hpp> so edecimal can use scan_decimal_float.
Test helpers and regression suite elastic/decimal/conversion/string_parse.cpp	Adds CheckParse/CheckReject helpers, reorganizes main into labeled test suites covering canonical integers, scientific-notation integers, allowed decimal-point forms (non-negative effective exponent), rejection of fractional/negative-effective-exponent inputs, malformed-token rejection, overly-large-expansion guard, negative-zero normalization, and operator>> behavior tests (failbit on bad token/empty stream, successful scientific parsing and consumption). Also standardizes exception reporting and final test reporting.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related issues

• feat(erational): extend parse to p/q form and decimal/scientific notation #855: erational::parse enhancement depends on expanded edecimal::parse support for decimal and scientific notation.

Possibly related PRs

• stillwater-sc/universal#838: Introduces scan_decimal_float tokenization primitive used by this PR.
• stillwater-sc/universal#858: Related expansion of the edecimal string-parse regression executable that this PR further develops.
• stillwater-sc/universal#840: Parallel refactor of scan_decimal_float tests using the same test-reporting pattern.

Suggested labels

enhancement

Poem

"I nibble digits, hop through dots and e,
From integers to exponents, parse sets me free.
No tiny fractions lost on my trail,
Zero sheds its sign without fail—🐰
Hooray for precise decimal tea!"

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 42.86% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately and concisely summarizes the main change: extending edecimal's parse function to handle decimal-point and scientific notation forms.
Linked Issues check	✅ Passed	All primary objectives from issue #854 are met: parse extended to accept integers, decimals, and scientific notation with effective exponent >= 0; scan_decimal_float reused for tokenization; negative fractional exponents rejected; operator>> hygiene added with failbit on parse failure; comprehensive regression tests added covering all specified cases.
Out of Scope Changes check	✅ Passed	All code changes are directly within scope: parse rewrite uses scan_decimal_float as specified, string_parse.hpp header included as required, test suite expanded to cover all acceptance criteria with no unrelated changes.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/issue-854-edecimal-parse-decimal-and-sci

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@include/sw/universal/number/edecimal/edecimal_impl.hpp`:
- Around line 393-405: The exponent expansion can grow eff_exp (computed from
scan.exp10 - scan.frac_part.size()) to an unbounded value and then push_back
zeros in a loop; guard this by validating eff_exp before any allocations: in the
code around eff_exp, add a check that eff_exp is non-negative and below a
configured safe limit (e.g. MAX_ALLOWED_EXPANSION or derived from a
MAX_TOTAL_DIGITS), and also ensure (scan.int_part.size() + scan.frac_part.size()
+ eff_exp) does not exceed a MAX_TOTAL_DIGITS cap; if the check fails, return
false (as done for negative) to avoid the clear()/push_back loops in push_back
and the trailing-zero loop, and consider using reserve/resize instead of
repeated push_back when within limits (referencing symbols eff_exp, scan.exp10,
scan.frac_part.size(), clear(), push_back).

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: feb64424-b290-4e02-b85e-69a5846843e5

📥 Commits

Reviewing files that changed from the base of the PR and between 842e381 and 4bd7897.

📒 Files selected for processing (3)

• elastic/decimal/conversion/string_parse.cpp
• include/sw/universal/number/edecimal/edecimal.hpp
• include/sw/universal/number/edecimal/edecimal_impl.hpp

[coveralls]

Coverage Report for CI Build 26009791606

Warning

Build has drifted: This PR's base is out of sync with its target branch, so coverage data may include unrelated changes.
Quick fix: rebase this PR. Learn more →

Coverage remained the same at 84.166%

Details

• Coverage remained the same as the base build.
• Patch coverage: 22 of 22 lines across 1 file are fully covered (100%).
• 10 coverage regressions across 1 file.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

10 previously-covered lines in 1 file lost coverage.

File	Lines Losing Coverage	Coverage
include/sw/universal/verification/test_suite_randoms.hpp	10	31.43%

---

Coverage Stats

Relevant Lines:	55519
Covered Lines:	46728
Line Coverage:	84.17%
Coverage Strength:	6419966.99 hits per line

---

💛 - Coveralls