Conversation
update with current main
- Add expand_metadata_columns() for expanding multiple columns efficiently - Add expand_single_metadata_column() for single column operations - Standardizes separator pattern to ',\\s*' for consistent handling - Reduces code duplication and improves maintainability
- Replace base R strsplit/table approach with tidyr-based helper - Standardizes separator pattern to ',\\s*' for consistent handling - More readable and maintainable code - Output format remains compatible with existing code
- Replace three separate separate_rows() calls with single expand_metadata_columns() call - More efficient single-pass expansion of all metadata columns - Standardizes separator pattern to ',\\s*' - Maintains same functionality and output format
- Replace three separate if blocks with a loop using column mapping - Use expand_single_metadata_column() instead of separate_rows() - Standardizes separator pattern to ',\\s*' - More maintainable and efficient code - Maintains all existing functionality including label warnings
- Update calculate_initial_records() to use expand_single_metadata_column() - Update calculate_detailed_records() to use expand_single_metadata_column() - Update calculate_phase_records() to use expand_single_metadata_column() and expand_metadata_columns() - Standardizes separator pattern to ',\\s*' throughout - More efficient and maintainable code
- Replace two separate_rows() calls with single expand_metadata_columns() call - Standardizes separator pattern to ',\\s*' - More efficient single-pass expansion - Removes redundant str_trim() calls (handled by helper function)
- Replace separate_rows() with expand_single_metadata_column() - Use proper column reference instead of positional index - Standardizes separator pattern to ',\\s*' - More maintainable and consistent code
- Fix incorrect calculation: use nrow(rv) instead of nrow(rv) - rv is expanded metadata (count_unique output) which inflates the count - rv is the actual unique citations after deduplication - Add calculation of duplicates removed (n_citations - n_unique_records) - Improve message formatting with: - Clear line breaks for readability - Number formatting with commas (e.g., 9,175 instead of 9175) - More informative structure showing totals, unique, and duplicates removed - Apply fixes to both app.R and app2.R - Resolves issue where message showed more unique citations than total citations
- Replace separate_rows() calls in unique_separated_phase() with expand_metadata_columns() - Replace separate_rows() + mutate(trimws) + filter in detailed_table_data() with expand_single_metadata_column() - Standardizes separator pattern to ',\\\\s*' throughout - More efficient and consistent with R package functions
- Replace separate_rows() calls in unique_separated_phase() with expand_metadata_columns() - Replace separate_rows() + mutate(trimws) + filter in detailed_table_data() with expand_single_metadata_column() - Replace separate_rows() in completeness_data() with expand_single_metadata_column() - Standardizes separator pattern to ',\\\\s*' throughout - More efficient and consistent with R package functions
- Update all calls to expand_metadata_columns() and expand_single_metadata_column() to use CiteSource::: prefix - This allows Shiny apps to access internal (non-exported) helper functions - Fixes error: 'could not find function expand_single_metadata_column' - Helper functions remain internal (not exported) as per best practices - Apply changes to both app.R and app2.R
- Exclude 'unknown' cite_source values after expanding metadata columns - Records with 'screened' or 'final' labels intentionally have empty cite_source - These get converted to 'unknown' during deduplication with show_unknown_tags=TRUE - Including 'unknown' in detailed record table causes: * Misleading source row that isn't actually a search source * Incorrect counts (Records Imported, Distinct Records) * Skewed percentage calculations (Source Contribution %, etc.) * Affected Total row calculations - Filtering 'unknown' ensures table only analyzes actual search sources - Consistent with calculate_phase_records() which already filters 'unknown' - Does not affect other tables/visuals (precision/sensitivity table uses different function) - Apply fix to both app.R and app2.R
- Added interactive "Use This" buttons to Card View for granular selection of Title, Author, Abstract, and Journal. - Implemented custom JavaScript to handle cross-column button toggling and provide immediate visual feedback. - Added server-side listener `input$field_preference_click` to store user preferences. - Implemented `apply_field_preferences` logic to overwrite surviving records with user selections in the final merged dataset. - Updated "Default" badge logic to correctly handle cases where one value is missing. - Fixed `record_id` vs `duplicate_id` column resolution during merge. - Resolved "unknown column" warnings by properly initializing `field_preferences` before JSON serialization.
- Add parameters use_custom, blocking_rounds, validation_criteria to dedup_citations(). - When use_custom=TRUE, call internal dedup_citations_custom() with configurable blocking rounds and validation criteria; fall back to default ASySD on error or if custom implementation is not loaded. - New R/dedup_custom.R: custom ASySD wrapper with default blocking rounds and validation criteria matching ASySD behaviour, plus optional stats on which criteria identified duplicate pairs. Shiny app (inst/shiny-app/CiteSource/app2.R): - Wire in custom deduplication option so users can run configurable dedup from the app (e.g. UI for use_custom and related options). Documentation: - Add AUTHOR_HANDLING.md: end-to-end documentation of author name handling, including expected format (e.g. "Last, First and Last, First"), cleaning rules, and behaviour in import, cleaning, deduplication, citation generation, export, and data conversion.
…nette revisions - Shiny app.R: full UI overhaul with bslib cards, workflow stepper, export hub, bidirectional filters, smart empty states, card view dedup, global.R bootstrap - Vendor ASySD deduplication engine into R/asys_dedup.R; remove ASySD GitHub dependency - Drop app2.R (superseded by refactored app.R) - Migrate all pipes to native |>; vectorize APA citation generation - CRAN compliance: specific @importFrom declarations, globalVariables, remove plogr - Add renv and renv.lock for reproducible dependency management - Revise vignettes: benchmark testing, screening phases, db-validation, topic coverage - Add search string comparison vignette - Fix deployment workflow: restore rsconnect + deploy steps, add RENV_CONFIG_SNAPSHOT_VALIDATE - Update CITATION.cff to v0.2.0 with release date - NEWS.md: full v0.2.0 changelog - Add custom_dedup_notes.md to .gitignore (local reference doc, not for commit)
- Remove citesource_working_example and citesource_benchmark_testing (deleted) - Add citesource_search_string_comparison (new) - Change url from http to https to match DESCRIPTION
- Remove deprecated record_counts_table(), record_summary_table(), precision_sensitivity_table() from R/tables.R - Add CITESOURCE2_ANALYSIS.md to .Rbuildignore - Fix pkgdown workflow: install from local source (devtools::install) instead of remotes::install_github() which installed from main branch
split(.$type) and split(.$facet) use magrittr's dot which is not supported by the native pipe - replace with anonymous function wrappers
calculate_initial_records(), calculate_detailed_records(), and calculate_phase_records() moved to count.R; create_initial_record_table(), create_detailed_record_table(), and create_precision_sensitivity_table() moved to tables.R. Delete new_count_and_table.R.
- Add CITATION.cff, REQUIREMENTS.*, ASySD info to .Rbuildignore - Document show_labels and log_scale params in plot_source_overlap_heatmap - Fix devtools::install(upgrade = 'never') to upgrade = FALSE (older devtools compat)
…urce pak (used by extra-packages) marks transitive deps as 'deps' source type, which shinyapps.io rejects. renv::install installs into renv's own library with proper GitHub source metadata that rsconnect and shinyapps.io both understand.
Use renv::install for CiteSource
Fix workflow conditional syntax for GitHub Pages deploy step
- Fix Title to use title case per CRAN policy - Rewrite Description to not start with package name - Add CRAN eval guard to all vignettes so chunks are skipped when vignette data is absent (R CMD check), preventing file-not-found errors - Exclude vignette data directories, shinytest fixtures, and .claude worktree from build via .Rbuildignore, reducing tarball from 31 MB to 2.6 MB
- Update.Rbuildignore to prevent worktree from entering tarball
- Quote 'RIS' and 'CSV' in Description to avoid spell-check flags
- Replace pre-built with prebuilt in Description to avoid hyphen-split flag
- Fix maintainer email to tnriley@gmail.com
- Replace \dontrun{} with \donttest{} in reimport_csv example
- Update cran-comments.md: correct note count, add reverse dependency statement
- Remove "An R Package for" from Title field per CRAN policy - Add single quotes around 'shiny' in Description field - Update RoxygenNote and add Config/roxygen2/version after roxygen2 upgrade
- Add @return/@value tags to all five flagged exported functions: export_bib, export_ris, plot_contributions, plot_source_overlap_upset, reimport_ris; side-effect functions use "No return value" phrasing - Remove default filename values from export_csv(), export_ris(), and export_bib() so functions no longer write to getwd() by default - Update examples to use tempfile() instead of bare filenames - Fix write_refs() fallback (file=TRUE) to write to tempdir() - Fix reimport_csv() example: replace \donttest{} with if(interactive()) to prevent execution of a non-existent path during --run-donttest check - Regenerate all affected .Rd files via devtools::document()
- Add CRAN-SUBMISSION to .Rbuildignore to suppress top-level file note - Add resubmission section to cran-comments.md with point-by-point responses to CRAN reviewer feedback - Update R CMD check results to reflect clean 0/0/1 result - Track CRAN-SUBMISSION artifact from initial submission
Replace .data$id with -"id" in across() call per tidyselect 1.2.0
Previously, confirming manual duplicate pairs triggered generate_dup_id and merge_metadata across all unique citations regardless of how many pairs were selected. For large datasets this was unnecessarily slow. Now only the records whose duplicate_id appears in additional_pairs are processed. Unaffected records are passed through unchanged and recombined with bind_rows, turning an O(N) operation into O(manual pairs).
Adds a `fields` parameter to `export_csv()` with three modes: - "full" (default): all columns, reimportable into CiteSource - "standard": core bibliographic + cite_source/label/string, suitable for import into RELApp and other screening tools - character vector: user-defined custom column selection Non-full exports emit a warning that the file cannot be reimported into CiteSource via reimport_csv(). In the Shiny app, adds a radio button preset selector and a conditional custom column picker to the CSV download card. Also fixes the Shiny CSV download handler to use export_csv() (which correctly sets row.names = FALSE) instead of write.csv().
…el position - Standard preset label now lists exact fields instead of vague description - Custom column picker marks required CiteSource reimport fields with ' *' - Adds legend note '* required for CiteSource reimport' in custom picker - Moves Cite CiteSource panel to top of export tab
Renames citesource_analysis_across_screening_phases.rmd to .Rmd so pkgdown finds it on case-sensitive Linux CI runners. Fixes broken expression syntax in manual deploy workflow condition.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.