Hubbry Logo
Sinhala input methodsSinhala input methodsMain
Open search
Sinhala input methods
Community hub
Sinhala input methods
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Sinhala input methods
Sinhala input methods
Sinhala input methods
View on Wikipedia
from Wikipedia
A screencast showing Sinhala text being entered using the Intelligent Input Bus into Firefox under Linux. As shown in the video, under the Singlish method, to input විකිපීඩියා (Wikipedia), wikipIdiyaa is entered.

Sinhala input methods are ways of writing the Sinhala language, spoken primarily in Sri Lanka, using a computer. Sinhala input methods can be broadly classified into two main groups: ones based on typewriter keyboard layouts, and ones that are meant to be typed on QWERTY keyboards using an input method, known as "Singlish".[1]

Wijesekara keyboard

[edit]

The Wijesekara keyboard is the standard typewriter keyboard for the Sinhala script. This keyboard layout was first created and approved by the government of Sri Lanka in 1964.

Windows Sinhala layout

In 2004, it was given the SLS standards as the Sri Lanka Sinhala Character Code for Information Interchange, SLS 1134 : 2004.[2]

Implementations

[edit]

The first standards compliant Sinhala Keyboard for Apple iOS was created by Bhagya Silva. This implementation featured a copyrighted custom layout that was based on SLS 1134:2004 optimised for mobile keyboards.[3]

Virtual Keyboards

[edit]

The first Sinhala virtual keyboard is "Helakuru". Helakuru was developed by Bhasha Lanka (Pvt) Ltd for Android and iOS devices. It was first released on Android in 2011 and in 2015 it was released on App Store also.[4] In 2019, Apple allowed Sinhala to be a keyboard layout and an iPhone language to boost Apple product sales in Sri Lanka.[citation needed]

References

[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Sinhala input methods

View on Grokipedia
from Grokipedia
Sinhala input methods are software and hardware techniques designed to facilitate the entry of text in the , an derived from the ancient and used primarily for the in . These methods map keystrokes or other inputs to the Sinhala block (U+0D80–U+0DFF), which encodes 18 independent vowels, 41 , dependent vowel signs, and special signs like the virama (al-lakuna) for consonant clusters. The core approach relies on sequential input of base followed by combining vowel signs and modifiers, with rendering engines handling visual reordering for elements like pre-base vowels or conjunct forms using the (U+200D). The development of Sinhala input methods began in the late 1980s amid efforts to enable Sinhala computing in , initially through non-standard encodings like SLASCII before transitioning to international standards. In 1996, the Standards Institution (SLSI) introduced SLS 1134 as the national standard for Sinhala character coding, revised in 2004 and 2011 to align with ISO/IEC 10646 and , specifying 128 code points for characters and keyboard sequences. This standard formalized the "type as you write" principle, where users input elements in phonetic or visual order, such as a consonant followed by al-lakuna for pure forms or kombuva for the 'e' sound. The Wijesekera keyboard layout, originally for typewriters and adapted for computers, became the basis, assigning up to 196 keys across shift states on a standard 101-key board. Traditional Sinhala input methods emphasize direct keyboard layouts, with the SLS 1134-compliant Wijesekera layout mapping Sinhala characters to keys via shift, AltGr, and join modifiers for conjuncts like rakaransaya (r-form) or yansaya (ya-phala). For instance, the letter ක (ka) is entered directly, while complex forms like ලක්ෂ (lakṣa) use between consonants. Phonetic layouts, such as those in Input Tools, allow from Romanized "" (e.g., "lansha" for ලංශ), converting inputs in real-time to Sinhala. Voice-based methods, supported by tools like Voice Typing, enable dictation in Sinhala for hands-free entry. Modern implementations extend these methods across platforms, with Windows supporting the standard Sinhala layout (introduced in Vista) and the Wijesekera variant (Wij 9) for precise character access since 2008. On mobile devices, Android and offer dedicated Sinhala keyboards via apps compliant with SLS 1134, including phonetic and gesture-based options, while on-screen keyboards provide fallback for desktops. The Information and Communication Technology Agency (ICTA) of promotes open-source tools through its Local Languages Portal, ensuring accessibility for Unicode fonts like Iskoola Pota. These methods address challenges like the script's 60+ conjunct possibilities and right-to-left vowel rendering, promoting digital inclusion for over 16 million Sinhala speakers.

Background

Overview of Sinhala Script

The is an , in which consonants carry an inherent sound that can be modified or suppressed using diacritic marks known as vowel signs or matras. It comprises 56 letters in total, consisting of 18 independent vowels and 38 consonants, with additional vowel signs serving as modifiers to alter the inherent vowel of a consonant or form standalone syllables. These elements allow for the representation of Sinhala's syllabic structure, where each akṣara (syllable) typically combines a consonant base with optional vowel modifications. The script derives from the ancient , specifically its southern variant used in early Sri Lankan inscriptions dating back to the 3rd century BCE, evolving through influences like the Pallava around the 4th century CE. Over time, it developed a distinctive rounded and style, characterized by circular letterforms adapted for writing on ola (palm leaves), which contrasts with the more angular shapes of northern Indic scripts like . This visual fluidity enhances readability on traditional materials while maintaining compatibility with modern digital rendering. Unique to Sinhala phonology, the script accommodates conjunct characters formed by combining consonants with a (halant) to suppress the inherent , creating clusters essential for loanwords from or . It also features prenasalization, where nasal sounds precede stops (e.g., /ᵐb/, /ⁿd/), represented by dedicated letters such as ඹ for mba and ධ for nda. , or the lengthening of for emphasis or grammatical distinction, is indicated by doubling consonants consecutively, underscoring the script's adaptation to Sinhala's prosodic features like consonant length contrasts.

Challenges in Digital Input

The Sinhala script's syllabic nature, derived from Brahmi, involves combining base consonants with vowel diacritics, conjuncts, and special marks such as the yansaya and rakaransaya, resulting in nearly 2,300 distinct glyphs that must be rendered accurately in digital environments. This complexity demands sophisticated rendering engines to handle ligatures and positional variants, often leading to input errors or incomplete character formation when users attempt to compose syllables in real-time on standard computing interfaces. Traditional Sinhala keyboard layouts, such as those mapping characters to a modified grid, deviate significantly from the familiar English arrangement, where keys like "" produce Sinhala consonants instead of Latin letters, creating a steep for bilingual users accustomed to English typing. This mismatch increases and error rates during between Sinhala and English, as typists must mentally remap key functions, hindering efficient input on shared devices or multilingual workflows. Prior to the adoption of in 1999, Sinhala lacked a universal encoding standard, relying on proprietary schemes like SLASCII (introduced in 1990), which utilized platform-specific fonts and 8-bit mappings that were incompatible across systems. Consequently, text files exchanged between different software or hardware often displayed as garbled symbols or unreadable characters, as the absence of the exact matching font rendered Sinhala glyphs as arbitrary Latin approximations or blank spaces. This fragmentation severely limited digital adoption of Sinhala in early eras, exacerbating accessibility barriers for non-English content.

Historical Development

Early Typewriter Layouts

The introduction of Sinhala typewriters in the early marked a significant advancement in mechanical input for the , largely driven by British colonial presses established in Ceylon to support administrative documentation, missionary publications, and local newspapers. These presses, building on earlier Dutch innovations from the 18th century, adapted European technology to the rounded and conjunct-heavy nature of Sinhala characters, enabling more efficient production of printed materials beyond traditional palm-leaf manuscripts. Early models focused on manual type composition before evolving into full mechanical typewriters by the and , primarily used in offices and printing houses. The Wijesekara layout emerged in as the first government-approved standardized typewriter keyboard for Sinhala, developed to address the inconsistencies in prior arrangements and facilitate widespread use in newspaper production and official typing. Named after its creator, S. Wijesekara, this layout was designed specifically for mechanical typewriters produced by companies like Olympia International, becoming the for Sinhala typists in . It prioritized ergonomic efficiency for professional users, such as journalists and clerks, by mapping characters based on of Sinhala text in contemporary publications. Key features of the Wijesekara layout included a strategic of consonants, vowels, and frequent conjuncts (such as prenasalized forms like ඟ and ඳ) to minimize keystrokes for common word formations, reflecting the script's inherent challenges with ligatures and diacritics. The keyboard incorporated dead keys—non-printing modifiers that combined with subsequent strikes to produce vowel signs (e.g., ා for the 'a' kara) and halant forms for conjuncts—allowing typists to generate over 50 basic characters and numerous combinations without dedicated keys for every variant. While early mechanical versions had approximately 44-50 keys, typical for compact designs, the layout's principles influenced later expansions, ensuring compatibility with the script's 60+ akṣara while avoiding excessive complexity. This design significantly boosted typing speeds in workflows, with adoption spreading rapidly among Sinhala newspapers by the late .

Transition to Digital Standards

In the late 1980s, the transition from mechanical typewriters to digital input for Sinhala began with the development of proprietary encodings tailored for early word processors in Sri Lanka. These systems emerged to address the limitations of typewriter-based layouts, which relied on physical typebars and were ill-suited for computer processing. One notable example was the Mahanama encoding, patented by Saputantri Mahanama and marketed through JRL in 1986, which enabled Sinhala text input on personal computers by mapping characters to custom code pages. This proprietary approach allowed for the creation of dedicated Sinhala word processors, such as those developed by local companies, but it resulted in fragmented compatibility across different software and hardware. The Computer and Information Technology Council of (CINTEC), established in , played a key role in these initiatives by forming the Committee on National Languages and IT (CANLIT) to define the Sinhala character set. During the , government initiatives spearheaded efforts to standardize digital Sinhala input and fonts to foster broader adoption. The Standards Institution (SLSI), under the Ministry of Science and Technology, recognized the need for consistent digital representation beyond proprietary systems. A key outcome was the establishment of SLS 1134 in 1996, which defined a Sinhala-specific encoding scheme inspired by the Indian ISCII model but adapted for Sinhala's unique conjuncts and diacritics. These initiatives, supported by CINTEC, promoted the creation of digital fonts and encouraged their integration into word processing applications, marking a shift toward government-backed . Initial attempts to extend ASCII for Sinhala input, such as the Sri Lanka Standard Code for Information Interchange (SLASCII) introduced in 1990, provided a foundation but highlighted significant limitations by the early 2000s. SLASCII extended the 7-bit ASCII framework to accommodate Sinhala characters within an 8-bit code page, enabling basic text handling in bilingual environments. However, its incompatibility with international systems, restricted character set (covering only modern Sinhala without full historical support), and issues with data portability across platforms fueled calls for global standards. By 2000, these shortcomings prompted advocacy from Sri Lankan technologists and institutions like CINTEC for inclusion in international frameworks, emphasizing the need for a universal encoding to support web and cross-border digital communication.

Types of Input Methods

Typewriter-Based Keyboards

Typewriter-based keyboards for Sinhala input replicate the fixed mapping of characters from traditional mechanical to digital keyboards, assigning specific keys to base , , and modifiers in a manner that follows the script's visual and structural composition rather than phonetic order. This approach employs a "type as you write" , where users input characters in the sequence they appear in or print—such as placing vowel signs above or below —using modifier keys or sequences to form conjuncts and ligatures. The layout requires memorization of non-phonetic positions, as keys are positioned based on historical and frequency of use in Sinhala text, often prioritizing common glyphs on easily accessible keys. These keyboards offer significant advantages for users trained on physical Sinhala typewriters, particularly professional typists in printing and publishing sectors, by providing a direct analog to familiar hardware that minimizes the and maintains high typing speeds for routine work. The fixed assignments ensure consistency across documents and devices, facilitating accurate reproduction of complex script elements like the al-lakuna ( omission mark) without additional software interpretation. However, disadvantages include reduced efficiency for novice users, as the arbitrary key placements demand extensive practice and can lead to errors in formation, especially for those accustomed to phonetic systems. In digital adaptations, typewriter-based layouts are implemented through custom keyboard drivers that emulate the original ergonomics while supporting encoding for cross-platform compatibility. On Windows, these are integrated via standard editors (IMEs) and drivers like those compliant with SLS 1134, allowing direct mapping of typewriter keys to code points (U+0D80–U+0DFF) for seamless text entry in applications. For , open-source configurations and frameworks such as SCIM or IBUS provide similar ports, using layout files to replicate the fixed assignments and handle modifier sequences for Sinhala glyphs.

Phonetic and Transliteration Methods

Phonetic and methods for Sinhala input allow users to type using the Latin alphabet on standard keyboards, with software automatically converting the input to Sinhala characters based on phonetic approximations or direct character mappings. In these schemes, known as "" or similar transliteration systems, users enter Romanized text that reflects the of Sinhala words, such as typing "siyalla" to produce සියල්ල (meaning "all"). This approach relies on real-time conversion engines that process the Latin input by , outputting the corresponding Sinhala akṣara (consonant-vowel combinations) without requiring users to memorize a specialized layout. Common mappings in these methods follow phonetic principles, where individual Latin letters or digraphs correspond to Sinhala graphemes. For instance, the vowel "a" maps to the independent vowel අ, while "aa" maps to the long vowel ආ; consonants like "k" map to the base ක (with inherent vowel), and combinations such as "ka" yield ක or "ki" yield කි by appending the appropriate vowel sign (pili). More complex examples include "sri" transliterating to ශ්‍රී (a honorific prefix) or "obata" to ඔබට (meaning "to you"), where the software handles conjuncts and diacritics automatically upon pressing the spacebar or . These mappings are designed to approximate , treating the script's nature—where consonants carry an inherent 'a' vowel modified by trailing matras—through sequential Latin keystrokes. These methods offer significant benefits for users proficient in English, particularly in bilingual environments or among the , as they enable seamless switching between Latin and Sinhala scripts on standard keyboards without additional hardware. Tools implementing these schemes, such as Input Tools or standalone converters, facilitate faster entry speeds for English-familiar typists compared to traditional layouts, with reported efficiencies in real-time applications like messaging and document . They also support on mobile and desktop platforms, allowing diaspora communities to maintain linguistic ties without learning complex key remappings. However, phonetic and methods face limitations due to the inherent ambiguities in Sinhala , where homophones—words with identical pronunciations but distinct orthographies—can lead to incorrect conversions without contextual disambiguation. For example, similar-sounding syllables might map to multiple possible graphemes, resulting in higher error rates in unpredicted inputs, especially for less common vocabulary or dialects. Additionally, the reliance on phonetic can struggle with Sinhala's script-specific features, such as retroflex sounds or distinctions, potentially requiring manual corrections and reducing overall accuracy in formal writing.

Predictive and Intelligent Input

Predictive and intelligent input methods for Sinhala leverage statistical language models and machine learning algorithms to anticipate and suggest text completions, enhancing typing efficiency beyond basic phonetic or transliteration mappings. These approaches typically process partial input—often in Romanized form corresponding to Sinhala phonetics—and generate predictions based on contextual probabilities derived from large corpora of Sinhala text. N-gram models, which estimate the likelihood of word sequences by analyzing contiguous n-word patterns (such as bigrams for two words or trigrams for three), form the foundation of many such systems. For instance, a composite n-gram model combining bigrams and trigrams has been applied to Sinhala corpora from newspapers, achieving up to 41% prediction accuracy in domain-specific contexts like sports articles. Machine learning techniques further refine these predictions by incorporating user patterns and error-prone elements unique to Sinhala, such as complex clusters and diacritics. In applications, a trigram-based model optimized with genetic algorithms personalizes suggestions according to time-series data and recipient categories, reducing average keystrokes per word from 15.45 in non-predictive entry to as low as 8.38 for frequent users—a savings of approximately 46%. Such models integrate into editors (IMEs) via dynamic dictionaries, where partial phonetic inputs trigger ranked word suggestions in pop-up menus, allowing selection with minimal additional keystrokes. This is particularly beneficial for Sinhala's script, where predicting full akṣara (syllabic units) from phonetic prefixes minimizes manual entry of diacritics like signs (e.g., ◌ා for '') or (◌්) for clusters. Recent advances as of 2025 incorporate neural architectures like BERT for reverse of Romanized Sinhala ("") to native script, improving handling of ambiguities through contextual embeddings and achieving higher accuracy in low-resource settings. approaches, adapting pre-trained models such as DeepSpeech to Sinhala corpora, further enhance predictive by fine-tuning on limited data, reducing errors in real-time input. Error correction features enhance intelligence by detecting and suggesting fixes for common input mistakes, such as incorrect placement or confusions arising from phonetic . Tools employing rule-based systems augmented with , like hybrid models, identify grammatical and errors in Sinhala sentences with 84.4% overall success, proposing corrections that align with syntactic patterns. For specifically, suggestion engines using edit-distance algorithms prioritize fixes for errors (e.g., mismatches) and achieve 62.3% accuracy on the first suggested correction, drawing from curated Sinhala dictionaries. When embedded in IME frameworks, these capabilities provide real-time feedback, such as auto-correcting omissions during phonetic typing, and can reduce keystrokes by 18-46% overall, with higher savings (up to 50%) for repetitive or domain-common terms like legal or political .

Voice Input Methods

Voice input methods for Sinhala primarily involve speech-to-text (STT) technologies designed to transcribe spoken Sinhala into its script, facilitating dictation without manual typing. These systems address the unique phonetic inventory of Sinhala, which includes approximately 40 phonemes—26 consonants and 14 vowels—along with features like prenasalization and that complicate accurate recognition. Sinhala's phonetic challenges, such as coarticulation in rapid speech and dialectal variations, necessitate tailored acoustic models to map audio inputs to textual outputs effectively. Core to these methods are acoustic models trained specifically on Sinhala phonemes, often employing architectures like deep neural networks (DNNs), time-delay neural networks (TDNNs), or end-to-end (E2E) models such as LF-MMI to capture nuances in and prosody. These models handle rapid speech by incorporating hidden Markov models (HMMs) or subspace Gaussian mixture models (SGMMs) for probabilistic alignment, with training data derived from low-resource corpora to mitigate data scarcity in Sinhala. Performance improves with techniques like speaker adaptation, which fine-tunes models to individual voices, reducing errors from accent or speed variations. For clear dictation, these systems achieve accuracy rates of around 85-95%, corresponding to word error rates (WER) as low as 11.72% in fine-tuned setups, with further gains from user-specific training that adapts models to personal speaking patterns. In command-based tasks, accuracies exceed 90%, demonstrating robustness for structured inputs. Recent developments as of 2025 include transfer learning with models like DeepSpeech for improved low-resource ASR, end-to-end speech-to-speech chatbots integrating ASR with natural language processing for interactive applications, and specialized tools for dyslexia screening and autism speech therapy, achieving enhanced accuracy through domain adaptation. Key applications encompass real-time transcription in mobile apps for tasks like and messaging, enabling seamless Sinhala input on smartphones. tools leverage these methods for visually impaired users, allowing voice dictation of documents and navigation in apps like Sinhala book readers. Additionally, (IVR) systems on mobile platforms use Sinhala STT for services such as banking queries and , promoting broader digital inclusion.

Specific Implementations

Wijesekara Keyboard

The Wijesekara keyboard, named after its designer, serves as the foundational standard for input on typewriters and computers. It was approved by the as the national Sinhala typewriter keyboard in 1968, marking a significant step in standardizing mechanical input for the language. This layout was later extended by the Computer and Information Technology Council of Sri Lanka (CINTEC) in the late for electronic typewriters and early personal computers, incorporating additional characters such as the "fa" (ෆ) to accommodate modern printing needs. By the 1970s, it had become the official layout for Sinhala printing in governmental publications and media outlets. The layout is designed to be compatible with the English keyboard base, facilitating bilingual use while dedicating specific rows to Sinhala characters. The top row primarily maps vowels and vowel signs, such as අ (a) on the W key, උ (u) on shift+W, and ඔ (o) on shift+T, following a phonetic order for ease of typing. The middle and bottom rows accommodate consonants, with common ones like ම (ma) on U, ස (sa) on I, and ද (da) on O positioned on the home row for frequent access; rarer conjuncts and aspirated forms, such as ඨ (ṭha) on G and ඪ (ḍa) on V, occupy the lower row. Modifiers for diacritics, ligatures, and special marks—like rakaransaya (r-vowel sign) on the % key, yansaya (independent ya) on H, and repaya (r-pe) on the —are clustered on the right side or accessed via Alt Gr combinations, enabling the formation of complex conjuncts such as ඳ (n̆da) via Alt Gr + O. This structure supports approximately 65 core Sinhala characters, with extensions for compliance allowing up to 2,300 glyphs through combining sequences. Standardization for digital use came with Sri Lanka Standard (SLS) 1134 in 1996, revised in 2001, 2004, and 2011 by the Standards Institution to align with ISO/IEC 10646 and , ensuring portability across computing platforms. The 2004 revision specifies keystrokes for 128 code points in the Sinhala (U+0D80–U+0DFF), including independent vowels, consonant-vowel combinations, and for halant forms. The 2011 revision (third revision) further refined compatibility with . In practice, the Wijesekara keyboard remains dominant in Sri Lankan government offices and media sectors, where it underpins official documents, election broadcasts (as seen in national TV displays since 1982), and printing presses. Key remapping software, such as implementations in Keyman for Windows and distributions like , allows users to toggle between English and Sinhala modes, supporting its integration into modern operating systems without hardware changes.

Sayura Scheme

The Sayura Scheme is a quasi- for the , developed by Anuradha Ratnaweera in mid-2004 specifically for environments to facilitate Sinhala text entry using standard keyboards. Unlike full phonetic transliteration systems that require spelling out words in Romanized form (e.g., "mama" for මම), Sayura employs direct Latin key mappings to Sinhala characters, emphasizing unmodified and context-dependent modifications to reduce typing overhead. This approach was initially implemented as a GTK-based module, later ported to Qt in September 2004 and to the SCIM framework in October 2005, with the name "Sayura" formalized in version 0.3.x released in May 2008. Key assignments in Sayura prioritize intuitive mappings for English typists, assigning single Latin letters to core Sinhala consonants while using vowel combinations for and long forms. For instance, the key "s" produces ස (sa), "y" produces ය (ya), "r" produces ර (ra), and "t" produces ත (ta), with uppercase variants for aspirated or special forms like "P" for ඵ (pha). follow a similar logic: "a" yields අ (a) or serves as a after consonants, "i" produces ඉ (i), and doubled keys like "aa" generate long vowels such as ආ (ā); additional combinations handle rarer forms, such as "kii" after a consonant for acute vowels (e.g., කී for ki). Context sensitivity ensures efficiency, where the same key like "i" adapts based on its position relative to consonants, minimizing the need for numerous dedicated keys and avoiding ambiguities inherent in Romanized systems. The philosophy behind Sayura focuses on minimal for users familiar with English keyboards, promoting direct character input over word-level prediction to enable straightforward web and document composition without extensive reconfiguration. It eschews full to prevent errors from Sinhala's phonetic inconsistencies, instead leveraging byte-based internal processing mapped to the Sinhala for compatibility. Adoption has been prominent in open-source ecosystems, serving as the primary Sinhala in / distributions; it was shipped with and systems, and provided installation packages via sinhala.sourceforge.net. Extensions include dedicated engines for frameworks like IBUS (ibus-sayura, version 1.3.2) and Fcitx (fcitx-sayura, version 0.1.2), making it a default option in until proposals in 2021 to transition to m17n variants, and it remains integrated in tools for online content creation across forums and blogging platforms.

Google Input Tools

Google Input Tools provides a transliteration-based for Sinhala, enabling users to type using Romanized "" (phonetic approximations in ) that is automatically converted to the . This fuzzy phonetic mapping allows for flexible input, such as typing "namaskara" to generate "නමස්කාර" or similar variants, with context-aware suggestions appearing as candidates for selection via keyboard shortcuts or clicks. The tool supports both online and offline modes; the online version operates via web interfaces like services, while offline functionality is available through downloadable installers for Windows and the Android keyboard app. In addition to , Google Input Tools integrates voice typing capabilities for Sinhala, primarily through and , where users can dictate text directly into documents or apps. Sinhala voice support was added in 2017 as part of an expansion to 30 new languages, enhancing accessibility for Sri Lankan users by converting spoken Sinhala into Unicode-compliant text with reasonable accuracy for standard dialects. This feature relies on Google's Cloud Speech-to-Text , which handles Sinhala (Sri Lanka variant) among over 120 supported languages. Customization options include a personal dictionary that learns from user corrections, allowing additions of proper names, uncommon terms, or preferred transliterations to improve future suggestions, particularly for diacritics and vowel signs in . For instance, repeated corrections for names like "" (කොළඹ) refine the tool's predictions over time. The system provides context-sensitive diacritic handling, suggesting appropriate conjuncts or matras based on preceding characters to reduce typing errors. The tool's global reach extends across platforms, including Chrome OS via browser extensions for seamless web input, Android devices through the integrated Gboard keyboard with transliteration and voice modes, and web-based access in services like Gmail and Google Translate. This multi-platform availability, updated periodically for better dialect recognition in Sri Lankan Sinhala, makes it a widely adopted solution for diaspora and local users without dedicated hardware keyboards.

Other Notable Methods

Ralla is an open-source transliteration input system for the , developed in 2021 specifically for the framework on operating systems. It enables users to type Sinhala text using Romanized "Singlish" input on standard keyboards, converting intuitive English-like phonetic spellings into in real-time. Based on the established Sayura scheme, Ralla optimizes for users by integrating seamlessly with IBUS and m17n libraries, allowing easy installation via package managers and configuration through input method switchers. Key features include support for vowels, consonants, modifiers, and , with mappings such as "a" to අ and "k" to ක, making it suitable for non-expert typists accustomed to informal texting. SriShell Primo, introduced in 2008, represents an early predictive phonetic tailored for efficient Sinhala text entry, particularly on resource-constrained devices like early mobile phones. This word-based system uses principled key assignments derived from phonetic transcriptions, where users enter Romanized sequences that closely match Sinhala pronunciation, such as "desei" for the word meaning "eyes." It employs a TRIE-structured compiled from approximately 240,000 words sourced from the Divaina corpus (January 2005 to May 2006), enabling dynamic of complete words or phrases as the user types, including support for compound words with omitted spaces. Designed for user-friendliness, SriShell achieves an average typing efficiency of 2.1 keystrokes per character by covering intuitive Roman input variations and reducing the need for exact character mappings. The method's phonetic foundation prioritizes flow over strict , enhancing speed and accuracy for mobile environments where full keyboards are unavailable.

Virtual and On-Screen Keyboards

Desktop Virtual Keyboards

Desktop virtual keyboards for Sinhala provide on-screen interfaces that enable users to input Sinhala characters directly on personal computers without relying on physical keyboards, particularly useful for systems lacking dedicated Sinhala hardware layouts. These tools typically display a grid of Sinhala glyphs and modifiers, allowing selection via mouse clicks or other pointing devices, and integrate with the operating system's framework to render Unicode-compliant text in applications. Prominent software examples include the built-in Windows Sinhala Keyboard, which leverages the On-Screen Keyboard (OSK) application to support Sinhala input layouts such as the standard typewriting arrangement. When the Sinhala language pack is installed via Windows Settings > Time & Language > Language, users can switch to it using Win + Space, enabling the OSK to display and input Sinhala characters interactively. Similarly, Ubuntu has offered built-in virtual keyboard support for Sinhala since version 9.10 (released in 2009, with enhancements by 2010), accessible through the Accessibility panel under Settings, where layouts like Sinhala phonetic or Wijesekara can be selected for on-screen use via tools such as the default screen keyboard or Onboard. Third-party solutions like Keyman Desktop further extend this by providing dedicated on-screen keyboards tailored for Sinhala, compatible with Windows, Linux, and macOS, and allowing layout switching within the application. These virtual keyboards feature clickable grids representing Sinhala consonants, vowels, and conjuncts, often with visual feedback for modifier states like vowel signs or halant marks upon hovering or selection. For enhanced usability on larger displays or touch-enabled desktops, options include zoom functionality to enlarge the keyboard grid—available in Windows OSK via window resizing and in Ubuntu's Onboard through scaling settings—and customization such as repositioning the keyboard or altering key sizes to suit user preferences. Such adaptations make them suitable for hybrid input on monitors. In terms of accessibility, desktop virtual keyboards for Sinhala support stylus input for precise character selection on compatible hardware, integrating seamlessly with mouse emulation for users with motor impairments. They also work with screen readers; for instance, Windows OSK is compatible with Narrator, which announces selected Sinhala characters aloud, while Ubuntu's implementation pairs with to provide auditory feedback for non-visual navigation and input in Sinhala text fields. These features ensure broader usability for disabled individuals typing in Sinhala on desktop environments.

Mobile Input Methods

Mobile input methods for Sinhala have evolved to accommodate touch-based interfaces on smartphones and tablets, focusing on phonetic and to handle the language's complex script with 41 consonants and numerous s. , Google's for Android, provides comprehensive Sinhala support, including native script typing and glide typing (swipe gestures) for efficient input. This feature allows users to trace letters across the keyboard to form words phonetically, such as swiping from 's' to 'i' to 'n' to 'h' for "සිංහ" (, meaning ), enhancing speed on small screens. Haptic feedback is configurable in for key presses and gesture completion, providing tactile confirmation during selection via long-press menus. Microsoft SwiftKey offers predictive Sinhala input primarily through English , where users type Romanized approximations like "api" to suggest "අපි" (api, meaning we), with the engine learning from usage for improved accuracy over time. This method supports up to five languages simultaneously on Android and up to two on , including Sinhala, and includes flow typing—a swipe similar to Gboard's—for faster phonetic entry. On , SwiftKey's Sinhala functionality is limited to using Romanized input (as of 2025), without a direct native script keyboard. Local developers have filled gaps with apps like Helakuru, a Sri Lankan that integrates phonetic and Wijesekara layouts for both Android and , emphasizing offline predictions and real-time suggestions tailored to colloquial Sinhala. Adoption of these mobile methods is driven by Android's dominance in , where over 80% of smartphones run the OS, enabling widespread use of and Helakuru for everyday tasks like messaging and social media. On , third-party keyboards such as Desh Sinhala Keyboard and Smart Sinhala Keyboard provide haptic-enabled with word predictions, addressing limitations in Apple's default Sinhala input, which lacks robust autocorrect. Gesture-based features, including swipe for multi-letter entry and haptic cues for precise placement, reduce errors in Sinhala's system, where vowel signs attach to consonants. Brief integration with voice input, as in 's 2017 addition of Sinhala speech recognition, further streamlines mobile composition for users on the go.

Standards and Unicode Support

Sinhala Unicode Block

The Sinhala Unicode block, designated as U+0D80–U+0DFF, encompasses 128 code points dedicated to encoding characters for the , primarily supporting the spoken in , as well as and texts written in this script. Introduced in version 3.0 in September 1999, the block initially allocated 80 code points, with subsequent versions expanding to 90 assigned characters by Unicode 6.0 and 91 as of Unicode 16.0 (2023), including independent vowels (e.g., U+0D85 SINHALA LETTER AYANNA), consonants (e.g., U+0D9A SINHALA LETTER ALPAPRAANA KAYANNA), dependent vowel signs (e.g., U+0DCF SINHALA VOWEL SIGN AELA-PILLA), and punctuation marks like U+0DF4 SINHALA PUNCTUATION KUNDDALIYA. The remaining code points are reserved for future allocation, ensuring extensibility while maintaining compatibility across digital systems. Composition of complex Sinhala forms relies on specific sequences to represent conjunct s and vowel matras, avoiding the need for precomposed glyphs. The (U+0DCA SINHALA SIGN AL-LAKUNA) suppresses the inherent vowel of a , and when followed by another , it forms a basic cluster; for explicit joining in ligatures or touching forms, the zero-width joiner (ZWJ, U+200D) is inserted after the , as in the sequence + + ZWJ + (e.g., U+0D9A U+0DCA U+200D U+0DDB for a joined "la" + "ra"). This approach, distinct from many Indic scripts, allows flexible rendering of al-lakuna () behaviors and supports modern Sinhala without ZWJ for standard stacked forms, while ZWJ enables optional touching or split matra representations. Rendering engines interpret these sequences to produce visually connected or stacked glyphs, ensuring accurate display of conjuncts like rakaaransaya or yansaya. Proper rendering of Sinhala text demands fonts with robust support to handle the script's characteristic rounded, circular —shaped by historical influences like palm-leaf inscriptions—which require precise positioning and substitution for aesthetic and readability consistency across platforms. Key features include 'akhn' for akhand ligatures, 'rphf' for repaya forms, 'vatu' for below-base substitutions like rakaaransaya, and 'pstf' for post-base splitting, implemented via GSUB ( substitution) and GPOS ( positioning) tables. These features enable shaping engines like Uniscribe or to reorder elements (e.g., placing pre-base vowels before the consonant) and apply contextual forms, preventing distortion of rounded contours in clusters or with diacritics. Without such support, may appear disjointed or incorrectly stacked, particularly on diverse operating systems.

Standardization Efforts

The Unicode Consortium added the Sinhala script to the Unicode Standard in version 3.0, released in September 1999, establishing a universal encoding framework for the script's characters in the range U+0D80 to U+0DFF. This inclusion aligned Sinhala with ISO/IEC 10646, the international standard for character encoding, ensuring compatibility across diverse computing systems. In Sri Lanka, the Information and Communication Technology Agency (ICTA) has led national standardization since taking over from the Computer and Information Technology Council (CINTEC) in 2005, focusing on promoting Unicode adoption and developing local standards for input methods. A key contribution was ICTA's 2010 proposal to the Unicode Consortium for encoding Sinhala numerals, including Lith Illakkam digits and archaic forms, which were incorporated into Unicode version 6.0 (2010) and subsequent updates, enhancing numerical representation in Sinhala text. The Standards Institution (SLSI) formalized the Standard Sinhala Computer Keyboard Layout through SLS 1134, first published in 1996, with revisions in 2004 (second revision) and 2011 (third revision), based on the Wijesekara scheme, to provide a consistent mapping of Sinhala characters to keys. SLSI also certifies keyboard layouts, fonts, and related software for compliance, supporting in local implementations. Complementing these, open-source font projects such as Google's Sans Sinhala, released as part of the Noto font family, offer comprehensive glyph coverage and rendering support for Sinhala input in cross-platform applications. These institutional and community-driven initiatives have enhanced software by reducing legacy encoding conflicts, enabling reliable Sinhala input and display in international tools like web browsers and operating systems.

References

  1. https://www.unicode.org/charts/PDF/U0D80.pdf
  2. http://www.unicode.org/wg2/docs/n2737.pdf
  3. https://www.unicode.org/L2/L2004/04131-sls1134.pdf
  4. https://www.language.lk/wp-content/uploads/2018/03/SLS-1134-2011.pdf
  5. https://r12a.github.io/scripts/sinh/si.html
  6. https://www.icta.lk/local-languages-initiative
  7. https://learn.microsoft.com/en-us/globalization/windows-keyboard-layouts
  8. https://www.unicode.org/versions/Unicode15.0.0/ch13.pdf
  9. http://www.sundaytimes.lk/201115/education/evolution-of-the-sinhala-script-422243.html
  10. https://www.researchgate.net/publication/364164816_Articulation_Timing_and_Orthographical_Representation_of_Consonant_Gemination_in_Sinhala
  11. https://icter.sljol.info/articles/10.4038/icter.v8i1.7162
  12. https://www.researchgate.net/publication/220494237_From_the_Past_to_the_Present_Evolution_of_Computing_in_the_Sinhala_Language
  13. https://www.keithv.com/pub/devsinhalese/ShyamEtAlCHI2013.pdf
  14. https://www.ft.lk/opinion/Introduction-to-Unicode-and-how-to-type-and-store-in-Sinhala-using-Unicode-fonts/14-693745
  15. https://www.sanog.org/resources/sanog40/SANOG40_Conference-S1-Harsha-Keynote.pdf
  16. https://nagaokaut.repo.nii.ac.jp/record/841/files/z315.pdf
  17. https://www.researchgate.net/figure/Wijesekera-Sinhala-typewriter-keyboard-layout-1964-Courtesy-Olympia-International-8_fig6_269288863
  18. https://opendl.ifip-tc6.org/db/conf/ifip3/histedu2008/NandasaraM08.pdf
  19. https://www.ict-history.lk/wp-content/uploads/LLBook.pdf
  20. https://www.language.lk/wp-content/uploads/2018/03/Recent-Developments-in-Local-Language-Computing-in-Sri-Lanka.pdf
  21. https://www.researchgate.net/publication/237385709_Development_of_Standards_for_Sinhala_Computing
  22. https://www.ft.lk/Opinion-and-Issues/Introduction-to-Unicode-and-how-to-type-and-store-in-Sinhala-using-Unicode-fonts/14-693745
  23. https://www.researchgate.net/publication/225627619_Sinhala_Computing_in_Early_Stage_-Sri_Lanka_Experience
  24. https://www.academia.edu/2818543/Challenges_of_enabling_IT_in_the_Sinhala_Language
  25. https://www.microsoft.com/globaldev/handson/dev/Unicode-KbdsonWindows.pdf
  26. https://www.unicode.org/standard/standard.html
  27. https://www.academia.edu/2818555/Development_of_Standards_for_Sinhala_Computing
  28. https://www.easysinhalatyping.com/
  29. https://accentgenerator.com/transliteration/english-to-sinhala-typing/
  30. https://helpcentre.lk/knowledgebase/different-input-methods-for-sinhala-part-i/
  31. https://arxiv.org/pdf/2404.13350
  32. http://ir.lib.ruh.ac.lk/xmlui/bitstream/handle/iruor/10495/ristcon_2017_proceedings-85.pdf?sequence=1&isAllowed=y
  33. https://www.ijsrp.org/research-paper-1213/ijsrp-p2473.pdf
  34. https://aclanthology.org/2025.indonlp-1.16.pdf
  35. https://www.researchgate.net/publication/377261579_Exploring_Model_Level_Transfer_Learning_For_Improving_Sinhala_Speech_Recognition
  36. https://www.researchgate.net/publication/349882648_Grammatical_error_detection_and_correction_model_for_Sinhala_language_sentences
  37. https://arxiv.org/pdf/2107.02983
  38. https://voxify.ai/blog/sinhala-text-to-speech
  39. https://josuk.sljol.info/articles/2739/files/submission/proof/2739-1-9532-1-10-20110124.pdf
  40. https://ieeexplore.ieee.org/document/9325468
  41. https://aclanthology.org/2021.icon-main.26.pdf
  42. https://www.researchgate.net/figure/System-architecture-of-Sinhala-speech-recognition-system_fig1_327517552
  43. https://www.semanticscholar.org/paper/Speaker-Adaptation-Applied-to-Sinhala-Speech-Nadungodage-Weerasinghe/cdc51799b6e47a85d3e9eaed1d756a29ed54a01e
  44. https://journals.sjp.ac.lk/index.php/vjs/article/view/6806/4830
  45. https://www.worldscientific.com/doi/10.1142/S2717554520500095
  46. https://dl.ucsc.cmb.ac.lk/jspui/bitstream/123456789/4905/1/20000715%252C%252020001207%252C%252020002009%2520.pdf
  47. https://acl-bg.org/proceedings/2025/RANLP%25202025/pdf/2025.ranlp-1.106.pdf
  48. https://ieeexplore.ieee.org/document/8421888/
  49. https://irjiet.com/common_src/article_file/1705755591_49f32d0b5c_8_irjiet.pdf
  50. https://github.com/FOC-SLIIT-Research-Project-2023/Mobile-Base-Sinhala-Book-Reader-for-The-Visually-Impaired-Individuals/releases
  51. https://www.researchgate.net/publication/327517552_Sinhala_Speech_Recognition_for_Interactive_Voice_Response_Systems_Accessed_Through_Mobile_Phones
  52. http://distro.ibiblio.org/hanthana/14.7/docs/Wijesekara-layout.pdf
  53. https://www.sayura.net/im/
  54. https://fedoraproject.org/wiki/Changes/ibus-m17n_as_default_Sinhala_IME
  55. https://packages.fedoraproject.org/pkgs/ibus-sayura/ibus-sayura/
  56. https://packages.debian.org/unstable/utils/fcitx-sayura
  57. https://www.google.com/inputtools/services/features/transliteration.html
  58. https://www.google.com/inputtools/try/
  59. https://www.google.com/intl/en-GB/inputtools
  60. https://blog.google/products/search/type-less-talk-more/
  61. https://www.dailymirror.lk/print/breaking-news/Google-adds-voice-recognition-for-Sinhala-and-Tamil/108-134707
  62. https://support.google.com/docs/answer/4492226?hl=en
  63. https://chromewebstore.google.com/detail/google-input-tools/mclkkofklkfljcocdinagocijmpgbhab?hl=en
  64. https://play.google.com/store/apps/details?id=com.google.android.inputmethod.latin
  65. https://suchthegeek.com/posts/2021/01/16/ralla-sinhala-transliterated-input-system-ibus
  66. https://aclanthology.org/I08-3009.pdf
  67. https://learn.microsoft.com/en-us/globalization/keyboards/kbdsn1
  68. https://support.microsoft.com/en-us/windows/use-the-on-screen-keyboard-osk-to-type-ecbb5e08-5b4e-d8c8-f794-81dbf896267a
  69. https://www.nongnu.org/sinhala/doc/howto/sinhala-howto.html
  70. https://help.ubuntu.com/stable/ubuntu-help/keyboard-osk.html.en
  71. https://keyman.com/keyboards/h/sinhala/
  72. https://support.google.com/gboard/answer/7068494
  73. https://support.google.com/gboard/answer/2811346
  74. https://support.google.com/gboard/answer/6102154
  75. https://support.microsoft.com/en-us/topic/what-languages-are-currently-supported-for-microsoft-swiftkey-keyboard-661bce6a-8446-435d-a6aa-9ea006ee8353
  76. https://support.microsoft.com/en-us/topic/how-to-use-microsoft-swiftkey-keyboard-with-more-than-one-language-33dc147c-9a6d-4016-bffd-339d8f50a6bb
  77. https://www.helakuru.lk/
  78. https://apps.apple.com/us/app/desh-sinhala-keyboard/id6449493844
  79. https://apps.apple.com/us/app/smart-sinhala-keyboard/id1445436313
  80. https://techcrunch.com/2017/08/14/googles-voice-typing-tech-adds-support-for-30-more-languages-reaching-further-into-africa-india/
  81. https://unicode-explorer.com/b/0D80
  82. https://www.compart.com/en/unicode/block/U%2B0D80
  83. https://learn.microsoft.com/en-us/typography/script-development/sinhala
  84. https://language.lk/wp-content/uploads/2017/11/Unicode-Fonts-Book.pdf
  85. https://www.ict-history.lk/en/language-technology/
  86. https://www.unicode.org/L2/L2010/10337-n3888-sinhala.pdf
Add your contribution
Related Hubs
User Avatar
No comments yet.