Code editors provide essential services that help developers understand, navigate, and modify programs. However, these services often fail in the presence of syntax errors. Existing syntax error recovery techniques, like panic mode and multi-option repairs, are either too coarse, e.g. in deleting large swathes of code, or lead to a proliferation of possible completions. This paper introduces tall tylr , an error-handling parser and editor generator that completes malformed code with syntactic obligations that abstract over many possible completions. These obligations generalize the familiar notion of holes in structure editors to cover missing operands, operators, delimiters, and sort transitions. tall tylr is backed by a novel theory of tile-based parsing, conceptually organized around a molder that turns tokens into tiles and a melder that completes and parses tiles into terms using an error-handling generalization of operator-precedence parsing. We formalize melding as a parsing calculus, meldr, that completes input tiles with additional obligations such that it can be parsed into a well-formed term, with success guaranteed over all inputs. We further describe how tall tylr implements molding and completionranking using the principle of minimizing obligations . Obligations offer a useful way to scaffold internal program representations, but in tall tylr we go further to investigate the potential of materializing these obligations visually to the programmer. We conduct a user study to evaluate the extent to which an editor like tall tylr that materializes syntactic obligations might be usable and useful, finding both points of positivity and interesting new avenues for future work.

Large language models (LLMs) have reshaped the landscape of program synthesis. However, contemporary LLM-based code completion systems often hallucinate broken code because they lack appropriate code context, particularly when working with definitions that are neither in the training data nor near the cursor. This paper demonstrates that tighter integration with the type and binding structure of the programming language in use, as exposed by its language server, can help address this contextualization problem in a token-efficient manner. In short, we contend that AIs need IDEs, too! In particular, we integrate LLM code generation into the Hazel live program sketching environment. The Hazel Language Server is able to identify the type and typing context of the hole that the programmer is filling, with Hazel’s total syntax and type error correction ensuring that a meaningful program sketch is available whenever the developer requests a completion. This allows the system to prompt the LLM with codebase-wide contextual information that is not lexically local to the cursor, nor necessarily in the same file, but that is likely to be semantically local to the developer’s goal. Completions synthesized by the LLM are then iteratively refined via further dialog with the language server, which provides error localization and error messages. To evaluate these techniques, we introduce MVUBench, a dataset of model-view-update (MVU) web applications with accompanying unit tests that have been written from scratch to avoid data contamination, and that can easily be ported to new languages because they do not have large external library dependencies. These applications serve as challenge problems due to their extensive reliance on application-specific data structures. Through an ablation study, we examine the impact of contextualization with type definitions, function headers, and errors messages, individually and in combination. We find that contextualization with type definitions is particularly impactful. After introducing our ideas in the context of Hazel, a low-resource language, we duplicate our techniques and port MVUBench to TypeScript in order to validate the applicability of these methods to higher-resource mainstream languages. Finally, we outline ChatLSP, a conservative extension to the Language Server Protocol (LSP) that language servers can implement to expose capabilities that AI code completion systems of various designs can use to incorporate static context when generating prompts for an LLM.

Type systems typically only define the conditions under which an expression is well-typed, leaving ill-typed expressions formally meaningless. This approach is insufficient as the basis for language servers driving modern programming environments, which are expected to recover from simultaneously localized errors and continue to provide a variety of downstream semantic services. This paper addresses this problem, contributing the first comprehensive formal account of total type error localization and recovery: the marked lambda calculus. In particular, we define a gradual type system for expressions with marked errors, which operate as non-empty holes, together with a total procedure for marking arbitrary unmarked expressions. We mechanize the metatheory of the marked lambda calculus in Agda and implement it, scaled up, as the new basis for Hazel, a full-scale live functional programming environment with, uniquely, no meaningless editor states. The marked lambda calculus is bidirectionally typed, so localization decisions are systematically predictable based on a local flow of typing information. Constraint-based type inference can bring more distant information to bear in discovering inconsistencies but this notoriously complicates error localization. We approach this problem by deploying constraint solving as a type-hole-filling layer atop this gradual bidirectionally typed core. Errors arising from inconsistent unification constraints are localized exclusively to type and expression holes, i.e., the system identifies unfillable holes using a system of traced provenances, rather than localized in an ad hoc manner to particular expressions. The user can then interactively shift these errors to particular downstream expressions by selecting from suggested partially consistent type hole fillings, which returns control back to the bidirectional system. We implement this type hole inference system in Hazel.

Structure editors have long promised to facilitate a continuous dialogue between programmer and system-one uninterrupted by syntax errors, such that vital program analyses and editor services are always available. Unfortunately, structure editors are notoriously slow or difficult to use, particularly when it comes to modifying existing code. Prior designs often struggle to resolve the tension between maintaining a program's hierarchical structure and supporting the editing affordances expected of its linear projection. We propose the paradigm of gradual structure editing, which mitigates this tension by allowing for temporary disassembly of hierarchical structures as needed for text-like editing, while scaffolding these interactions by generating syntactic obligations that, once discharged, guarantee proper reassembly. This paper contributes the design and evaluation of a gradual structure editor called teen tylr. We conducted a lab study comparing teen tylr to a text editor and a traditional structure editor on structurally complex program editing tasks, and found that teen tylr helped resolve most usability problems we identified in prior work, though not all, while achieving competitive performance with text editing on most tasks. We conclude with a discussion of teen tylr's remaining limitations and design implications for future code editors and parsers.

Programming has become a collaboration between human programmers, who drive intent, and interactive assistants that suggest contextually relevant editor actions. There has been considerable work on suggestion synthesis strategies—from semantic autocomplete to modern program synthesis, repair, and machine learning research. This diversity of contextually viable strategies creates a need for an integrative, human-centered perspective on the problem of programming assistant design that (1) confronts the problem of integrating a variety of synthesis strategies, fed by shared semantic analyses capable of operating on program sketches, and (2) centers the needs of the human programmer: comprehending, comparing, ranking, and filtering suggestions generated by various synthesizers, and in some cases participating in a synthesizer’s search by supplying additional expressions of intent. This paper contributes a conceptual architecture and API to guide programming assistant designers as they confront these integration and human-centered design challenges. We then instantiate this architecture with two prototype end-to-end assistant designs, both developed for the Hazel programming environment, that emphasize understudied design aspects, namely continuity, explainability, human-in-the-loop synthesis, and the integration of multiple analyses with multiple synthesis strategies.

Text editing is powerful, but some types of expressions are more naturally represented and manipulated graphically. Examples include expressions that compute colors, music, animations, tabular data, plots, diagrams, and other domain-specific data structures. This paper introduces live literals, or livelits, which allow clients to fill holes of types like these by directly manipulating a user-defined GUI embedded persistently into code. Uniquely, livelits are compositional: a livelit GUI can itself embed spliced expressions, which are typed, lexically scoped, and can in turn embed other livelits. Livelits are also uniquely live: a livelit can provide continuous feedback about the run-time implications of the client’s choices even when splices mention bound variables, because the system continuously gathers closures associated with the hole that the livelit is filling. We integrate livelits into Hazel, a live hole-driven programming environment, and describe case studies that exercise these novel capabilities. We then define a simply typed livelit calculus, which specifies how livelits operate as live graphical macros. The metatheory of macro expansion has been mechanized in Agda.