mdbplyr 0.4.1

This is a bug-fix release. All four issues were found by running mdbplyr against a live MongoDB server rather than only against mocked collections, and each is covered by a new regression test.

Bug fixes

• Fixed filter() with multiple predicates (e.g. filter(cond1, cond2)) compiling $and as a JSON object with numeric-string keys instead of an array, which MongoDB silently treated as a no-op filter.
• Fixed is.na() compiling to a bare $eq against null, which does not match a genuinely missing field under MongoDB’s $expr semantics; it now compiles through $ifNull first.
• Fixed comparisons (<, <=, >, >=, !=), !, and if_else() matching rows with a missing operand, because BSON sort order places missing/null below every number. Comparisons involving a field now compile with an NA-propagation guard so they drop or return NA for missing values the same way dplyr does.
• Fixed unwind_array() and flatten_fields() rejecting a nested array root (e.g. message.measurements) that infer_schema() only ever registers through its scalar leaves (message.measurements.Fx and so on). Both now resolve such paths directly, matching select()’s existing support for nested root paths.

mdbplyr 0.4.0

This release expands the lazy MongoDB translation layer introduced in v0.3.0, moving mdbplyr from a small core verb subset to a broader analytical backend with joins, window functions, richer summaries, and schema-aware selection helpers.

New verbs and query translations

• Added inner_join(), left_join(), semi_join(), and anti_join() methods for tbl_mongo.
• Joins compile to MongoDB $lookup pipelines. Mutating joins flatten matches by default with $lookup, $unwind, and $replaceRoot; unnest = FALSE keeps matches as a nested array column.
• Filtering joins keep the left-hand columns and filter rows by whether a match exists.
• Join support is intentionally conservative: the right-hand side must be a plain tbl_mongo collection reference with a known schema in the same database.

Expanded select() and expression support

• select() now supports name-based tidyselect helpers when the schema is known: starts_with(), ends_with(), contains(), matches(), everything(), all_of(), any_of(), ranges, and negation.
• select() still rejects where() because column types are unknown without reading data.
• coalesce() is now translated in scalar expressions and compiles to nested MongoDB $ifNull expressions.
• row_number() is available as an alias for the existing 1:n() row-numbering idiom in mutate() and transmute().

Grouping and summaries

• group_by() now supports named computed grouping keys, such as group_by(bucket = floor(amount / 10)).
• summarise() gained sd(), var(), first(), last(), and n_distinct().
• median() and quantile() are supported on MongoDB 7.0 or newer through MongoDB’s percentile accumulator.
• Summary expressions now use the same field-vs-local name resolution rules as other translated expressions.

across()

• Added across() support in mutate() and summarise().
• Supported selections are name-based and schema-aware, matching the new select() helper support.
• Supported functions are bare function names and ~ lambdas, with optional .names templates.
• across() deliberately does not support where(), anonymous functions, or functions stored in variables.

Window functions

• Added window function translation in mutate() and transmute() via MongoDB $setWindowFields.
• Supported ranking functions are rank(), min_rank(), and dense_rank().
• Supported cumulative functions are cumsum(), cummean(), cummax(), and cummin().
• Supported offset functions are lag() and lead().
• Window functions require MongoDB 5.0 or newer. Ranking functions sort by their column argument; cumulative and offset windows take their order from a preceding arrange().

Server version awareness

• mongo_src() and tbl_mongo() now accept an optional server_version argument.
• When possible, mongo_src() probes the connected MongoDB server with buildInfo.
• Added mongo_server_version() for mongo_src and tbl_mongo objects.
• Version-gated features now fail clearly on servers that are known to be too old.

Behavior fixes and internals

• mutate() assignments now compile in user order, so later expressions can refer to fields created earlier in the same call.
• Fixed a select() regression involving nested root paths.
• Pipeline compilation now preserves operation order more explicitly through the internal query representation, which is needed for ordered mutation, row-numbering, window functions, and joins.
• Added extensive mock-backed tests for the new phases and for flat, nested, and missing-field semantics.

Documentation and packaging

• The README and vignette now document the v0.4.0 feature set, limitations, and translation model.
• Installation instructions were updated for CRAN availability while preserving GitHub installation instructions for development builds.
• Generated vignette artifacts were removed from the source repository.