import { IFileInfo } from "@/FileSystem/Types"; import { EasyParserParent } from "./EasyParserParent"; import { IAtom } from "@/UI/Navigation/TreeView/TreeInterfaces"; import { GLModel } from "@/UI/Panels/Viewer/GLModelType"; interface ISDFBond { atomIndex1: number; // 0-indexed based on order in SDF atom block atomIndex2: number; // 0-indexed based on order in SDF atom block bondType: number; } /** * A parser for SDF (Structure-Data File) files. * Parses atom and bond information from the first molecule in an SDF string. * Supports both V2000 and V3000 formats. */ export class EasyParserSDF extends EasyParserParent { private _parsedBonds: ISDFBond[] = []; // Maps 0-based index in this._atoms to original 1-based index from SDF private _atomOrderToSdfIndex: Map = new Map(); // Maps original 1-based SDF index to 0-based index in this._atoms private _sdfIndexToAtomOrder: Map = new Map(); private _isV3000: boolean; constructor(src: IFileInfo | GLModel | IAtom[]) { super(undefined); this._isV3000 = (src as IFileInfo).contents.includes("V3000"); this._load(src); } /** * Load the source. * * @param {IFileInfo | GLModel | IAtom[]} src The source to parse. */ _load(src: IFileInfo | GLModel | IAtom[]): void { // Ensure fields are initialized before doing anything else if (!this._parsedBonds) { this._parsedBonds = []; } if (!this._atomOrderToSdfIndex) { this._atomOrderToSdfIndex = new Map(); } if (!this._sdfIndexToAtomOrder) { this._sdfIndexToAtomOrder = new Map(); } const lines = (src as IFileInfo).contents.split(/\r?\n/); this._atoms = []; this._parsedBonds = []; this._atomOrderToSdfIndex.clear(); this._sdfIndexToAtomOrder.clear(); if (this._isV3000) { this._parseV3000(lines); } else { this._parseV2000(lines); } } /** * Parses a V3000 formatted SDF file from an array of lines. * * @param {string[]} lines The lines of the SDF file. * @private */ private _parseV3000(lines: string[]): void { let inAtomBlock = false; let inBondBlock = false; let atomOrder = 0; for (const line of lines) { if (line.startsWith("M V30 END ATOM")) { inAtomBlock = false; continue; } if (line.startsWith("M V30 END BOND")) { inBondBlock = false; continue; } if (inAtomBlock) { this._atoms.push(line); // V3000 atom lines: M V30 [properties...] const parts = line.trim().match(/\S+/g); // Using match for robustness. if (parts && parts.length >= 3) { const sdfIndex = parseInt(parts[2], 10); if (!isNaN(sdfIndex)) { this._atomOrderToSdfIndex.set(atomOrder, sdfIndex); this._sdfIndexToAtomOrder.set(sdfIndex, atomOrder); atomOrder++; } } } else if (inBondBlock) { // V3000 bond lines: M V30 [properties...] const parts = line.trim().match(/\S+/g); // Using match for robustness. if (parts && parts.length >= 6) { const bondType = parseInt(parts[3], 10); const sdfAtomIdx1 = parseInt(parts[4], 10); const sdfAtomIdx2 = parseInt(parts[5], 10); const atomOrderIdx1 = this._sdfIndexToAtomOrder.get(sdfAtomIdx1); const atomOrderIdx2 = this._sdfIndexToAtomOrder.get(sdfAtomIdx2); if ( atomOrderIdx1 !== undefined && atomOrderIdx2 !== undefined && !isNaN(bondType) ) { this._parsedBonds.push({ atomIndex1: atomOrderIdx1, atomIndex2: atomOrderIdx2, bondType: bondType, }); } } } if (line.startsWith("M V30 BEGIN ATOM")) { inAtomBlock = true; } if (line.startsWith("M V30 BEGIN BOND")) { inBondBlock = true; } } } /** * Parses a V2000 formatted SDF file from an array of lines. * * @param {string[]} lines The lines of the SDF file. * @private */ private _parseV2000(lines: string[]): void { let atomCount = 0; let bondCount = 0; let countsLineIndex = -1; const currentAtomBlockLines: string[] = []; // Find the counts line (typically 4th line, 0-indexed) for the first molecule for (let i = 0; i < lines.length; i++) { const line = lines[i]; if (i >= 3 && line.trim().endsWith("V2000")) { const parts = line.trim().split(/\s+/); if (parts.length >= 2) { const numAtoms = parseInt(parts[0], 10); const numBonds = parseInt(parts[1], 10); if (!isNaN(numAtoms) && !isNaN(numBonds)) { atomCount = numAtoms; bondCount = numBonds; countsLineIndex = i; break; } } } if (line.trim() === "$$$$") { // End of the first molecule block break; } } if (countsLineIndex === -1) { // Attempt a heuristic parse for atoms if no counts line is found, this is less reliable let inMoleculeBlock = false; for (let i = 0; i < lines.length; ++i) { const line = lines[i]; if (line.includes("M END") || line.trim() === "$$$$") { break; } // Heuristic: Atom lines have X, Y, Z coords and an element symbol. // Example: " -0.6664 -0.2064 0.1642 C 0 0 0 0 0 0 0 0 0 0 0 0" if ( line.match( /^\s*-?\d+\.\d{4}\s+-?\d+\.\d{4}\s+-?\d+\.\d{4}\s+\w/ ) ) { // Assume past header if (!inMoleculeBlock && i > 2) { inMoleculeBlock = true; } if (inMoleculeBlock) { currentAtomBlockLines.push(line); } } else if ( inMoleculeBlock && currentAtomBlockLines.length > 0 && !line.match(/^\s*\d+\s+\d+\s+\d+/) ) { // If we were reading atoms and the next line doesn't look like a bond or atom, stop. break; } } this._atoms = currentAtomBlockLines; // Note: Bond parsing is skipped in this heuristic fallback as it's unreliable without counts. return; } // Extract atom lines const atomBlockStartIndex = countsLineIndex + 1; for (let i = 0; i < atomCount; i++) { const atomLineIndex = atomBlockStartIndex + i; if (atomLineIndex < lines.length) { currentAtomBlockLines.push(lines[atomLineIndex]); // Map the 0-indexed position in our _atoms array to the original 1-based SDF atom index this._atomOrderToSdfIndex.set(i, i + 1); } else { // console.warn("SDF parsing: Atom count in counts line exceeds available lines in file."); break; } } this._atoms = currentAtomBlockLines; // Extract bond lines const bondBlockStartIndex = atomBlockStartIndex + atomCount; for (let i = 0; i < bondCount; i++) { const bondLineIndex = bondBlockStartIndex + i; if (bondLineIndex < lines.length) { const line = lines[bondLineIndex]; // SDF bond line: aaabbbttt... (atom1_idx, atom2_idx, bond_type) // Atom indices are 1-based in SDF format. const atomIdx1 = parseInt(line.substring(0, 3).trim(), 10) - 1; // Convert to 0-indexed const atomIdx2 = parseInt(line.substring(3, 6).trim(), 10) - 1; // Convert to 0-indexed const bondType = parseInt(line.substring(6, 9).trim(), 10); if (!isNaN(atomIdx1) && !isNaN(atomIdx2) && !isNaN(bondType)) { this._parsedBonds.push({ atomIndex1: atomIdx1, atomIndex2: atomIdx2, bondType, }); } } else { // console.warn("SDF parsing: Bond count in counts line exceeds available lines in file."); break; } } } /** * Parse an atom line string from SDF into an IAtom object. This method is * called by the `getAtom` method of the parent class. * * @param {string} atomStr The atom line string. * @param {number} [atomParserIndex] The 0-based index of this atom in the * parser's `_atoms` list. This parameter * is required by EasyParserSDF. * @returns {IAtom | undefined} The parsed atom, or undefined if parsing * fails. */ _parseAtomStr( atomStr: string, atomParserIndex?: number ): IAtom | undefined { if (this._isV3000) { return this._parseV3000AtomStr(atomStr, atomParserIndex); } else { return this._parseV2000AtomStr(atomStr, atomParserIndex); } } /** * Parses a V3000 atom line string into an IAtom object. * * @param {string} atomStr The V3000 atom line string. * @param {number} [atomParserIndex] The 0-based index of this atom. * @returns {IAtom | undefined} The parsed atom. * @private */ private _parseV3000AtomStr( atomStr: string, atomParserIndex?: number ): IAtom | undefined { // M V30 [key=val] ... const parts = atomStr.trim().match(/\S+/g); if (!parts || parts.length < 7) return undefined; const sdfIndex = parseInt(parts[2], 10); const elem = parts[3]; const x = parseFloat(parts[4]); const y = parseFloat(parts[5]); const z = parseFloat(parts[6]); // parts[7] is mass difference, which we can ignore for now let charge = 0; // Look for charge property in remaining parts for (let i = 8; i < parts.length; i++) { if (parts[i].startsWith("CHG=")) { charge = parseInt(parts[i].substring(4), 10); if (isNaN(charge)) charge = 0; break; } } if (isNaN(x) || isNaN(y) || isNaN(z)) return undefined; const bonds: number[] = []; const bondOrder: number[] = []; if (atomParserIndex !== undefined) { this._parsedBonds.forEach((bond) => { if (bond.atomIndex1 === atomParserIndex) { bonds.push(bond.atomIndex2); bondOrder.push(bond.bondType); } else if (bond.atomIndex2 === atomParserIndex) { bonds.push(bond.atomIndex1); bondOrder.push(bond.bondType); } }); } return { x, y, z, elem, serial: sdfIndex, atom: elem, resn: "UNL", chain: "A", resi: 1, altLoc: " ", b: charge, bonds, bondOrder, }; } /** * Parses a V2000 atom line string into an IAtom object. * * @param {string} atomStr The V2000 atom line string. * @param {number} [atomParserIndex] The 0-based index of this atom. * @returns {IAtom | undefined} The parsed atom. * @private */ private _parseV2000AtomStr( atomStr: string, atomParserIndex?: number ): IAtom | undefined { // SDF atom line format (V2000): // Columns Content // 1-10 x-coordinate // 11-20 y-coordinate // 21-30 z-coordinate // 32-34 Atom symbol (left-justified) // 35-36 Mass difference (dd) from default isotope mass. (+/- relative to default) // 37-39 Charge (ccc): 0 (uncharged or not specified), 1 (+3), 2 (+2), 3 (+1), 4 (doublet radical), 5 (-1), 6 (-2), 7 (-3) // ... more fields exist but are not parsed by this basic parser for now. // Basic check if (atomStr.length < 34) { return undefined; } // atomParserIndex is essential for SDF bond parsing, ensure it's provided. if (atomParserIndex === undefined) { // This case should ideally not be reached if getAtom in parent always passes it. // console.error("EasyParserSDF._parseAtomStr: atomParserIndex is undefined. This is required for SDF parsing."); return undefined; } const x = parseFloat(atomStr.substring(0, 10).trim()); const y = parseFloat(atomStr.substring(10, 20).trim()); const z = parseFloat(atomStr.substring(20, 30).trim()); const elem = atomStr.substring(31, 34).trim(); // Atom symbol let charge = 0; if (atomStr.length >= 39) { const chargeCode = parseInt(atomStr.substring(36, 39).trim(), 10); switch (chargeCode) { case 1: charge = 3; break; case 2: charge = 2; break; case 3: charge = 1; break; // case 4: // doublet radical, charge is still 0. We don't store radical info in IAtom. case 5: charge = -1; break; case 6: charge = -2; break; case 7: charge = -3; break; default: charge = 0; break; // Includes 0 for uncharged } } const bonds: number[] = []; const bondOrder: number[] = []; // atomParserIndex is the 0-based index of the current atom line in this._atoms this._parsedBonds.forEach((bond) => { if (bond.atomIndex1 === atomParserIndex) { bonds.push(bond.atomIndex2); bondOrder.push(bond.bondType); } else if (bond.atomIndex2 === atomParserIndex) { bonds.push(bond.atomIndex1); bondOrder.push(bond.bondType); } }); const originalSdfAtomNumber = this._atomOrderToSdfIndex.get(atomParserIndex); return { x, y, z, elem, serial: originalSdfAtomNumber, // Use original 1-based SDF atom number for serial atom: elem, // In SDF, atom name is usually the element symbol resn: "UNL", // SDF doesn't typically have residue names chain: "A", // SDF doesn't typically have chains resi: 1, // SDF doesn't typically have residue indices altLoc: " ", // SDF doesn't have alternate locations in the same way as PDB b: charge, // Using 'b' for charge for consistency with EasyParserMol2 bonds, bondOrder, // hetflag, etc. are not typically part of basic SDF atom lines }; } /** * Get the atom at the given 0-based index. * Overrides parent to pass the index to `_parseAtomStr` for context. * * @param {number} idx The 0-based index of the atom in the internal list. * @returns {IAtom} The parsed atom object. * @throws {Error} if parsing fails. */ getAtom(idx: number): IAtom { const atomEntry = this._atoms[idx]; // If it's not a string, it's already been parsed and cached. if (typeof atomEntry !== "string") { return atomEntry as IAtom; } // Parse the string representation of the atom. const parsedAtom = this._parseAtomStr(atomEntry as string, idx); if (parsedAtom === undefined) { throw new Error( `Failed to parse SDF atom line at index ${idx}: "${atomEntry}"` ); } // Cache the parsed atom object to avoid re-parsing. this._atoms[idx] = parsedAtom; return parsedAtom; } }