Ch 07a : Recoding and verification

Record Owner/Creator

Checks that the owner and “Added by” references identify an existing user or group.

Essential for invalid owners: ownership controls editing permissions, and an invalid owner can make normal management of records difficult. An invalid creator mainly damages attribution. The repair assigns invalid references to Database Manager, ID 2.

Pointer Targets

Finds record pointers referring to records that do not exist.

Essential: these links cannot be followed, and connected searches, reports and relationship displays may be incomplete. The automatic repair deletes the faulty pointers; it cannot recover the intended target.

Target Types

Checks that a pointer’s target belongs to one of the record types permitted for that field.

Important: a link may open successfully but connect the wrong kind of information—for example, an Author field pointing to a Place. Searches and reports that assume the expected type can give misleading results.

Invalid Parents

Checks that parent–child connections agree in both directions: the parent identifies the child, and the child identifies the parent.

Important where parent–child structures are used: records may be omitted from child lists or appear inconsistently in navigation and editing. Some missing reverse pointers can be added automatically; other discrepancies require inspection.

Invalid/Duplicate Terms

Finds references to missing parent or inverse terms, and flags duplicate labels or labels ending in numbers that may have resulted from duplicate avoidance.

Important for broken term references: vocabulary hierarchy or inverse relationships may behave incorrectly. Usually lower priority for duplicate labels: these can split classification and search results, but identical labels in different contexts may be legitimate. Review before merging or removing terms.

Field Types

Checks field definitions for references to nonexistent terms, nonexistent non-selectable terms, or nonexistent record types in pointer constraints.

Usually a routine repair: stale references can affect available choices or constraints. Heurist offers Auto Repair. Repeated recurrence deserves investigation.

Default Values

Checks that defaults specified in record type definitions are valid for their fields.

Important for subsequent data entry: new records may receive an invalid default or fail to receive the intended value. Existing records are not necessarily affected.

Definitions Groups

Checks that record types, base fields and vocabularies belong to existing organisational groups.

Mostly a definition-management problem: definitions may be difficult to find in the design interface. The checker automatically places affected definitions in the appropriate Trash group when that group exists; this does not delete them.

Title Masks

Checks whether each record type has a valid expression for constructing record titles.

Important for usability: generated titles may fail or be incomplete, making records difficult to identify in results and pointer selectors. The underlying field values are not necessarily damaged.

Term Values

Finds term fields containing identifiers for terms that do not exist.

Essential for affected classifications: the value cannot be resolved to a valid term, compromising display and searching. The automatic repair deletes the faulty term values; it does not reconstruct their meaning.

Expected Terms

Checks that existing terms belong to the vocabulary or permitted term selection specified for the field.

Important: a recognisable term can still be inappropriate for its field, causing inconsistent classification and filtering. Where the permitted vocabulary contains the same label, Heurist can offer a replacement.

Empty Fields

Finds stored field entries containing null or empty values.

Generally minor housekeeping: an empty entry adds no useful information and can complicate value counts or processing. Removing it does not supply any missing information.

Non-Standard Fields

Finds values in fields that are not included in the record type’s defined field list.

Often harmless and sometimes intentional, particularly after imports or structural changes. The values remain stored, but may be overlooked during normal editing or reporting. Consider adding the field to the record type rather than deleting its data.

Required Fields

Finds required fields that are missing or empty.

Depends on the field: omission of an essential identifier or descriptive value may compromise the record; another required field may merely represent a completeness policy. The database can still operate, but the records do not meet their declared requirements.

Single Value Fields

Finds fields defined to accept one value that contain multiple values.

Important when the values conflict: different displays or processing routines may use one value or show several. Review the values and either retain the correct one or change the field’s repeatability.

Multiple Workflow Stages

Finds multiple workflow-stage values where the field permits only one.

Usually moderate: contradictory stages make workflow filtering unreliable. The checker automatically reduces them to one stage, preferring a non-importing stage where available.

Geo Values

Checks for missing or invalid geometry, inconsistent geometry descriptions, and coordinates outside valid bounds.

Important for mapping and spatial analysis: affected records may be absent from maps or appear in incorrect locations. Certain wrapped longitudes can be corrected automatically; other problems need manual repair.

Date Values

Checks whether date values can be recognised and interpreted, offering corrections for some formats and identifying values requiring manual attention.

Important for chronological work: invalid or misinterpreted dates can give incorrect date searches, sorting or timeline positions. Review ambiguous day/month interpretations carefully.

Spaces in Values

Checks text for leading and trailing whitespace, double spaces and longer sequences of whitespace.

Usually very minor: mainly affects presentation, although whitespace can interfere with matching or duplicate detection. Some whitespace is corrected automatically; longer sequences are offered for selected repair. Preserve intentional formatting in transcriptions or quotations.

Invalid Characters

Checks free-text and block-text values for characters regarded as invalid by the checker.

Variable importance: these can cause display or export problems. Inspect the reported characters before changing them, particularly in multilingual text or source transcriptions.

Date Index

Checks consistency between stored date fields and the supporting date index.

Important for reliable date searches: records may be missed or incorrectly selected even when their displayed dates look correct. Rebuilding the index repairs the supporting search data; invalid source dates require separate correction.

Relationship Cache

Checks the supporting cache used to represent record relationships.

Important for connected searches and relationship displays: links may be missing or inconsistent despite the underlying records being correct. The cache can be rebuilt from the stored data.

Check files for consistency

Finds missing, duplicate and unused uploaded files. Missing originals are serious: attachments cannot be opened and must be recovered or relinked. Duplicate or unused files are generally storage housekeeping; inspect them before removal.

Check URLs in text fields and titles

Checks record URLs and links within text fields. Broken links reduce access to referenced resources, but do not normally damage the database itself. A failed URL check can also reflect temporary unavailability or access restrictions.