data/src/FileSystem/LoadSaveMolModels/ParseMolModels/ParseMolecularModels.worker.ts

import {
    IMolData,
    sendResponseToMainThread,
    waitForDataFromMainThread,
} from "@/Core/WebWorkers/WorkerHelper";

import {
    IAtom,
    TreeNodeType,
    SelectedType,
} from "@/UI/Navigation/TreeView/TreeInterfaces";
import { randomID } from "@/Core/Utils/MiscUtils";
import {
    defaultProteinStyle,
    defaultNucleicStyle,
    defaultLigandsStyle,
    defaultMetalsStyle,
    defaultLipidStyle,
    defaultIonsStyle,
    defaultSolventStyle,
} from "@/Core/Styling/SelAndStyleDefinitions";
import {
    ionSel,
    lipidSel,
    metalSel,
    nucleicSel,
    proteinSel,
    solventSel,
    standardProteinResidues,
} from "../Types/ComponentSelections";
import { IFormatInfo, getFormatInfoGivenType } from "../Types/MolFormats";
import { getFileNameParts } from "@/FileSystem/FilenameManipulation";
import { TreeNode } from "@/TreeNodes/TreeNode/TreeNode";
import { TreeNodeList } from "@/TreeNodes/TreeNodeList/TreeNodeList";
import { IFileInfo } from "@/FileSystem/Types";
import { makeEasyParser } from "./EasyParser";
import { EasyParserParent } from "./EasyParser/EasyParserParent";
import { convertIAtomsToIFileInfoPDB } from "../ConvertMolModels/_ConvertIAtoms";
import { ISelAndStyle } from "@/Core/Styling/SelAndStyleInterfaces";
import { organizeNodesIntoHierarchy } from "@/UI/Navigation/TreeView/TreeUtils";

enum NodesOrModel {
    Nodes,
    Model,
}

const COLLAPSE_ONE_NODE_LEVELS = false;
const COVALENT_RADII: { [key: string]: number } = {
    H: 0.31, HE: 0.28, LI: 1.28, BE: 0.96, B: 0.84, C: 0.76,
    N: 0.71, O: 0.66, F: 0.57, NE: 0.58, NA: 1.66, MG: 1.41,
    AL: 1.21, SI: 1.11, P: 1.07, S: 1.05, CL: 1.02, K: 2.03,
    CA: 1.76, SC: 1.57, TI: 1.48, V: 1.44, CR: 1.39, MN: 1.39,
    FE: 1.32, CO: 1.26, NI: 1.24, CU: 1.32, ZN: 1.22, GA: 1.22,
    GE: 1.20, AS: 1.19, SE: 1.20, BR: 1.20, KR: 1.16, RB: 2.20,
    SR: 1.95, Y: 1.90, ZR: 1.75, NB: 1.64, MO: 1.54, TC: 1.47,
    RU: 1.46, RH: 1.42, PD: 1.39, AG: 1.45, CD: 1.44, IN: 1.42,
    SN: 1.39, SB: 1.39, TE: 1.38, I: 1.39, XE: 1.40, CS: 2.44,
    BA: 2.15, LA: 2.07, CE: 2.04, PR: 2.03, ND: 2.01, PM: 1.99,
    SM: 1.98, EU: 1.98, GD: 1.96, TB: 1.94, DY: 1.92, HO: 1.92,
    ER: 1.89, TM: 1.90, YB: 1.87, LU: 1.87, HF: 1.75, TA: 1.70,
    W: 1.62, RE: 1.51, OS: 1.44, IR: 1.41, PT: 1.36, AU: 1.36,
    HG: 1.32, TL: 1.45, PB: 1.46, BI: 1.48, PO: 1.40, AT: 1.50,
    RN: 1.50, FR: 2.60, RA: 2.21, AC: 2.15, TH: 2.06, PA: 2.00,
    U: 1.96, NP: 1.90, PU: 1.87, AM: 1.80, CM: 1.69
};
const DEFAULT_COVALENT_RADIUS = 0.76; // Approx Carbon
const BONDING_TOLERANCE = 0.45;
/**
 * Gets the covalent radius of an atom.
 *
 * @param {IAtom} atom The atom.
 * @returns {number} The radius.
 */
function getAtomRadius(atom: IAtom): number {
    const elem = atom.elem ? atom.elem.toUpperCase() : "";
    return COVALENT_RADII[elem] || DEFAULT_COVALENT_RADIUS;
}
/**
 * Checks if two groups of atoms are bonded.
 *
 * @param {EasyParserParent} group1 The first group.
 * @param {EasyParserParent} group2 The second group.
 * @returns {boolean} True if bonded.
 */
function areGroupsBonded(group1: EasyParserParent, group2: EasyParserParent): boolean {
    // Optimization: Check bounding boxes with a generous cutoff first
    // Max bond dist roughly 2.5A (S-S + tolerance).
    if (!group1.isWithinDistance(group2, 3.0)) {
        return false;
    }
    for (let i = 0; i < group1.length; i++) {
        const a1 = group1.getAtom(i);
        if (a1.x === undefined || a1.y === undefined || a1.z === undefined) continue;
        const r1 = getAtomRadius(a1);
        for (let j = 0; j < group2.length; j++) {
            const a2 = group2.getAtom(j);
            if (a2.x === undefined || a2.y === undefined || a2.z === undefined) continue;
            const r2 = getAtomRadius(a2);
            const distSq = (a1.x - a2.x) ** 2 + (a1.y - a2.y) ** 2 + (a1.z - a2.z) ** 2;
            const threshold = r1 + r2 + BONDING_TOLERANCE;
            if (distSq <= threshold * threshold) {
                return true;
            }
        }
    }
    return false;
}
/**
 * Helper function to generate default mol container (used throughout).
 *
 * @param  {string}       molName                            The name of the
 *                                                           container.
 * @param  {NodesOrModel} [nodesOrModel=NodesOrModel.NODES]  Whether to generate
 *                                                           nodes or model.
 * @returns {TreeNode}  The default mol container.
 */
function _getDefaultTreeNode(
    molName: string,
    nodesOrModel = NodesOrModel.Nodes
): TreeNode {
    let obj = {
        title: molName,
        viewerDirty: true,
        treeExpanded: false,
        visible: true,
        focused: false,
        selected: SelectedType.False,
    };

    obj = {
        ...obj,
        ...(nodesOrModel === NodesOrModel.Model
            ? { model: [] }
            : { nodes: new TreeNodeList() }),
    };

    return new TreeNode(obj);
}

/**
 * Divides a molecule into chains.
 *
 * @param  {any}     sel           The selection to divide (e.g., a selection
 *                                 that gets all protein atoms).
 * @param  {EasyParserParent} mol  The molecule with the atoms. Note that the
 *                                 atoms specified by sel end up getting
 *                                 removed, because the function moves them into
 *                                 their own molecules.
 * @param  {string}  molName       The name of the entry.
 * @returns {TreeNode} The molecule with the chains.
 */
function organizeSelByChain(
    sel: any,
    mol: EasyParserParent,
    molName: string
): TreeNode {
    let selectedAtoms = mol.selectedAtoms(sel, true);

    // If chain is " " for any atom, set it to "X"
    selectedAtoms = selectedAtoms.map((atom: IAtom) => {
        if (atom.chain === " ") {
            atom.chain = "X";
        }
        return atom;
    });

    const treeNode = _getDefaultTreeNode(molName);
    let lastChainID = "";
    selectedAtoms.forEach((atom: IAtom) => {
        const nodeList = treeNode.nodes as TreeNodeList;
        if (atom.chain !== lastChainID) {
            nodeList.push(
                new TreeNode({
                    title: atom.chain,
                    model: [],
                    viewerDirty: true,
                    treeExpanded: false,
                    visible: true,
                    selected: SelectedType.False,
                    focused: false,
                })
            );
            lastChainID = atom.chain;
        }

        (nodeList.get(nodeList.length - 1).model as IAtom[]).push(atom);
    });
    // mol.removeAtoms(selectedAtoms);

    return treeNode;
}

/**
 * Some molecular components don't need chains (e.g., solvents and ions). This
 * function flattens chains.
 *
 * @param  {TreeNode} treeNode The molecule (with chains) to flatten.
 * @returns {TreeNode} The flattened molecule.
 */
function flattenChains(treeNode: TreeNode): TreeNode {
    if (!treeNode.nodes) {
        throw new Error("No nodes found in treeNode.");
    }

    const flattened = _getDefaultTreeNode(treeNode.title, NodesOrModel.Model);

    treeNode.nodes.forEach((chain: TreeNode) => {
        if (!chain.model) {
            throw new Error("No atoms found in chain.");
        }

        (flattened.model as IAtom[]).push(...(chain.model as IAtom[]));
    });
    return flattened;
}

/**
 * Gets an id of a given atom (string representation).
 *
 * @param  {IAtom} atom The atom to get the id of.
 * @returns {string} The id of the atom.
 */
function residueID(atom: IAtom): string {
    return atom.resn + ":" + atom.resi;
}

/**
 * In some cases, it's useful to further divide chains into residues (e.g.,
 * small-molecule compounds).
 *
 * @param  {TreeNode} treeNode The molecule to divide.
 * @returns {TreeNode} The divided molecule.
 */
function divideChainsIntoResidues(treeNode: TreeNode): TreeNode {
    if (!treeNode.nodes) {
        return treeNode;
    }

    const dividedMolEntry = _getDefaultTreeNode(treeNode.title);

    let lastChainID = "";
    treeNode.nodes.forEach((chain: TreeNode) => {
        if (!chain.model) {
            // Already divided apparently.
            return;
        }

        if (chain.title === "" || chain.title === undefined) {
            // Default to chain A if not specified
            chain.title = "A";
        }

        if (chain.title !== lastChainID) {
            dividedMolEntry.nodes?.push(_getDefaultTreeNode(chain.title));
            lastChainID = chain.title;
        }

        let lastResidueID = "";
        (chain.model as IAtom[]).forEach((atom: IAtom) => {
            const chains = dividedMolEntry.nodes;
            if (!chains) {
                throw new Error("No chains found in dividedMolEntry.");
            }
            const residues = chains.get(chains.length - 1).nodes;
            if (!residues) {
                // Always exists. This here for typechecker.
                throw new Error("No residues found in dividedMolEntry.");
            }

            const newKey = residueID(atom);
            if (newKey !== lastResidueID) {
                residues.push(_getDefaultTreeNode(newKey, NodesOrModel.Model));
                lastResidueID = newKey;
            }
            const atoms = residues.get(residues.length - 1).model as IAtom[];
            if (atoms) {
                atoms.push(atom);
            }
        });
    });
    return dividedMolEntry;
}

/**
 * If any molecule has a list of 1 submolecules, collapse it so one molecule,
 * merging the titles.
 *
 * @param  {TreeNode} treeNode            The molecule to collapse.
 * @param  {boolean}       [childTitleFirst=false] When creating the merged
 *                                                 title, but the name of the
 *                                                 child molecule first.
 * @returns {TreeNode} The collapsed molecule.
 */
function collapseSingles(
    treeNode: TreeNode,
    childTitleFirst = false
): TreeNode {
    if (treeNode.nodes) {
        let anyNodeMerged = true;
        while (anyNodeMerged) {
            anyNodeMerged = false;

            const allNodes = new TreeNodeList([treeNode]).flattened;
            allNodes.forEach((anyNode: TreeNode) => {
                const anyNodeNodes = anyNode.nodes as TreeNodeList;
                if (anyNode.nodes && anyNodeNodes.length === 1) {
                    // 1 child node
                    const childNode = anyNodeNodes.get(0);
                    anyNode.title = childTitleFirst
                        ? `${childNode.title}:${anyNode.title}`
                        : `${anyNode.title}:${childNode.title}`;
                    anyNode.model = childNode.model;
                    anyNode.type = childNode.type;

                    if (childNode.nodes) {
                        anyNode.nodes = childNode.nodes;
                    } else {
                        delete anyNode.nodes;
                    }

                    anyNodeMerged = true;
                }
            });
        }
    }

    // Now for some purly cosmetic changes to the top-level menu items. (test
    // 1HU4, 1XDN, and 2HU4).

    if (treeNode.title.endsWith(":Compound")) {
        if (!treeNode.nodes || treeNode.nodes.length === 0) {
            // It has no children. Remove the "Compound" suffix.
            treeNode.title = treeNode.title.substring(
                0,
                treeNode.title.length - 9
            );
        } else {
            // It has children. Probably looks like "X:Compound"
            treeNode.title = treeNode.title.split(":")[1];
        }
    } else {
        // Looks like "Protein:A". Remove chain.
        treeNode.title = treeNode.title.split(":")[0];
    }

    return treeNode;
}

/**
 * Adds the molecule type.
 *
 * @param  {TreeNode}  treeNode  The molecule to add the type and style to.
 * @param  {ISelAndStyle[]} stylesAndSels The styles and selections to add.
 */
function addMolTypeAndStyle(treeNode: TreeNode, stylesAndSels: ISelAndStyle[]) {
    const molType = treeNode.type;
    new TreeNodeList([treeNode]).filters.onlyTerminal.forEach(
        (mol: TreeNode) => {
            mol.type = molType;
            mol.styles = stylesAndSels;
        }
    );

    new TreeNodeList([treeNode]).flattened.forEach((mol: TreeNode) => {
        mol.id = randomID();
        mol.treeExpanded = false;
        mol.visible = true;
        mol.viewerDirty = true;
    });
}

/**
 * Given molecular data, returns information about the format.
 *
 * @param  {IMolData} data  The molecular data.
 * @returns {IFormatInfo}  Information about the format.
 */
function getFormatInfo(data: IMolData): IFormatInfo {
    const molFormat = data.format;
    return getFormatInfoGivenType(molFormat) as IFormatInfo;
}

/**
 * Given molecular text, divides the text by frames.
 *
 * @param  {string} molText  The molecular text.
 * @param  {IFormatInfo} molFormatInfo  Information about the format.
 * @returns {string[]}  The frames.
 */
function divideMolTxtIntoFrames(
    molText: string,
    molFormatInfo: IFormatInfo
): string[] {
    let frames: string[] = [molText];

    // NOTE: If the file has been loaded through OpenBabel, it's already been
    // separated. But to keep the code organized, let's just process all
    // molecules the same way.

    if (molFormatInfo && molFormatInfo.frameSeparators) {
        for (const frameSeparator of molFormatInfo.frameSeparators) {
            // const txt = frameSeparator.text.replace(/\$/g, "\\$");
            if (frameSeparator.isAtEndOfFrame) {
                // Add string "<<DIVIDE>>" after every occurance of
                // molInfo.frameSeparators.text
                // const srch = new RegExp(txt + "$", "gm");
                // const rpl = frameSeparator.text + "<<DIVIDE>>";
                molText = molText.replaceAll(
                    frameSeparator.text,
                    // Strange exception required for SDF files
                    (frameSeparator.text !== "\n$$$$\n"
                        ? frameSeparator.text
                        : "\n$$$$$$$$\n") + "<<DIVIDE>>"
                );

                // Every time "\n$$$$\n" appears in string, replace with "MOOSE\n$$$$\n". Don't use regex / RegExp
            } else {
                // Add string "<<DIVIDE>>" before every occurance of
                // molInfo.frameSeparators.text
                molText = molText.replaceAll(
                    frameSeparator.text,
                    "<<DIVIDE>>" + frameSeparator.text
                );
            }
        }

        frames = molText.split("<<DIVIDE>>");
        frames = frames.map((f) => f.trim()).filter((f) => f.length > 0);
    }

    return frames;
}

/**
 * Given molecular text, tries to detect the name from the text itself. TODO:
 * Not currently using this function, but leave it if you need it.
 *
 * @param  {string} molText  The molecular text.
 * @param  {IFormatInfo} molFormatInfo  Information about the format.
 * @returns {string}  The name. Returns "" if no name found.
 */
function getNameFromContent(
    molText: string,
    molFormatInfo: IFormatInfo
): string {
    const regexps = molFormatInfo.extractMolNameRegex;
    let firstMatch = "";
    if (regexps) {
        for (const regexp of regexps) {
            // Must reset regex for repeated use. Interesting. See
            // https://stackoverflow.com/questions/4724701/regexp-exec-returns-null-sporadically
            regexp.lastIndex = 0;

            const match = regexp.exec(molText);
            if (match) {
                firstMatch = match[1];
                break;

                // below prevents multiple matches
                // molText = molText.replaceAll(match[1], "");
            }
        }
    }

    // Remove terminal ; from match
    if (firstMatch.endsWith(";")) {
        firstMatch = firstMatch.substring(0, firstMatch.length - 1);
    }

    // Keep only unique
    // allMatches = allMatches.filter((v, i, a) => a.indexOf(v) === i);

    // Keep only at most first 10 letters. Append ... if more than 10
    // letters.
    if (firstMatch.length > 20) {
        firstMatch = firstMatch.substring(0, 20) + "...";
    }

    // Compile into single string. Separate by ;
    return firstMatch.replaceAll(":", "");
}

/**
 * Given molecular data from the main thread, convert it into a TreeNode object
 * divided by component (protien, compound, solvent, etc.).
 *
 * @param  {IMolData} data The molecular data.
 * @returns {Promise<TreeNode>} The divided molecule.
 */
function dividePDBAtomsIntoDistinctComponents(
    data: IMolData
): Promise<TreeNodeList> {
    // Any molecules that share bonds are the same component.

    // Get the format
    const molFormatInfo = getFormatInfo(data);
    const frames = divideMolTxtIntoFrames(
        data.fileInfo.contents,
        molFormatInfo
    );

    const molName = getFileNameParts(data.fileInfo.name).basename;
    const frameTitles = frames.map((f, i) => {
        return `${getNameFromContent(f, molFormatInfo)}:${molName}:${i + 1}`;
    });

    const fileContentsAllFrames: TreeNodeList = new TreeNodeList();

    for (let frameIdx = 0; frameIdx < frames.length; frameIdx++) {
        const frame = frames[frameIdx];
        const frameTitle = frameTitles[frameIdx];
        const molWithAtomsToDivide = makeEasyParser({
            contents: frame,
            name: `tmp.${data.format}`,
        } as IFileInfo);

        if (molWithAtomsToDivide.length === 0) {
            // No atoms in model. Skip.
            continue;
        }
        // Create a flat list of terminal nodes
        const terminalNodes = new TreeNodeList();
        const componentSels: [any, TreeNodeType, boolean][] = [
            [proteinSel, TreeNodeType.Protein, false],
            [nucleicSel, TreeNodeType.Nucleic, false],
            [solventSel, TreeNodeType.Solvent, true],
            [metalSel, TreeNodeType.Metal, true],
            [ionSel, TreeNodeType.Ions, true],
            [lipidSel, TreeNodeType.Lipid, false],
        ];
        for (const [sel, type, flatten] of componentSels) {
            let selToUse = {};
            if (sel.or) {
                // Special case for ionSel
                for (const s of sel.or) {
                    selToUse = { ...selToUse, ...s };
                }
            } else {
                selToUse = sel;
            }
            const atoms = molWithAtomsToDivide.selectedAtoms(selToUse, true);
            if (atoms.length > 0) {
                if (flatten) {
                    terminalNodes.push(
                        new TreeNode({
                            title: type,
                            type: type,
                            model: atoms,
                            visible: true,
                            focused: false,
                            viewerDirty: true,
                            treeExpanded: false,
                            selected: SelectedType.False,
                        })
                    );
                } else {
                    // Group by chain
                    const atomsByChain: { [key: string]: IAtom[] } = {};
                    atoms.forEach((atom) => {
                        const chain = atom.chain || "A";
                        if (!atomsByChain[chain]) atomsByChain[chain] = [];
                        atomsByChain[chain].push(atom);
                    });
                    for (const chainId in atomsByChain) {
                        terminalNodes.push(
                            new TreeNode({
                                title: chainId,
                                type: type,
                                model: atomsByChain[chainId],
                                visible: true,
                                focused: false,
                                viewerDirty: true,
                                treeExpanded: false,
                                selected: SelectedType.False,
                            })
                        );
                    }
                }
            }
        }
        // Everything else is a compound, grouped by residue
        const remainingAtoms = molWithAtomsToDivide.atoms;
        if (remainingAtoms.length > 0) {
            // 1. Group by residue ID
            const atomsByResidue: { [key: string]: IAtom[] } = {};
            const resKeys: string[] = [];
            remainingAtoms.forEach((atom) => {
                const resId = `${atom.resn}:${atom.resi}:${atom.chain || "A"}`;
                if (!atomsByResidue[resId]) {
                    atomsByResidue[resId] = [];
                    resKeys.push(resId);
                }
                atomsByResidue[resId].push(atom);
            });
            // 2. Create parsers for each group to use optimization features
            const groups = resKeys.map((key) => ({
                key,
                atoms: atomsByResidue[key],
                parser: makeEasyParser(atomsByResidue[key]),
                merged: false,
                connectedTo: [] as number[], // indices of other groups
            }));
            // 3. Build connectivity graph optimized with spatial hashing
            const GRID_SIZE = 5.0; // Angstroms
            const CUTOFF = 3.0; // Matches areGroupsBonded check distance threshold
            const grid = new Map<string, number[]>();

            // Add groups to grid based on their bounding box
            groups.forEach((g, idx) => {
                const bounds = g.parser.getBounds();
                if (!bounds) return;
                const margin = CUTOFF / 2; // Expand bounds by half the cutoff on each side
                const iMin = Math.floor((bounds.minX - margin) / GRID_SIZE);
                const iMax = Math.floor((bounds.maxX + margin) / GRID_SIZE);
                const jMin = Math.floor((bounds.minY - margin) / GRID_SIZE);
                const jMax = Math.floor((bounds.maxY + margin) / GRID_SIZE);
                const kMin = Math.floor((bounds.minZ - margin) / GRID_SIZE);
                const kMax = Math.floor((bounds.maxZ + margin) / GRID_SIZE);
                for (let i = iMin; i <= iMax; i++) {
                    for (let j = jMin; j <= jMax; j++) {
                        for (let k = kMin; k <= kMax; k++) {
                            const key = `${i},${j},${k}`;
                            if (!grid.has(key)) grid.set(key, []);
                            grid.get(key)!.push(idx);
                        }
                    }
                }
            });
            // Identify potential pairs from the grid
            const potentialPairs = new Set<string>();
            for (const indices of grid.values()) {
                if (indices.length < 2) continue;
                for (let i = 0; i < indices.length; i++) {
                    for (let j = i + 1; j < indices.length; j++) {
                        const idx1 = indices[i];
                        const idx2 = indices[j];
                        // Canonical key to avoid duplicates (smaller index first)
                        const key = idx1 < idx2 ? `${idx1},${idx2}` : `${idx2},${idx1}`;
                        potentialPairs.add(key);
                    }
                }
            }
            // Check bonds only for spatially adjacent groups
            for (const pairKey of potentialPairs) {
                const [s1, s2] = pairKey.split(",");
                const i = parseInt(s1);
                const j = parseInt(s2);
                if (areGroupsBonded(groups[i].parser, groups[j].parser)) {
                    groups[i].connectedTo.push(j);
                    groups[j].connectedTo.push(i);
                }
            }
            // 4. Traverse graph to find connected components
            for (let i = 0; i < groups.length; i++) {
                if (groups[i].merged) continue;
                const componentIndices: number[] = [i];
                groups[i].merged = true;
                const stack = [i];
                while (stack.length > 0) {
                    const curr = stack.pop()!;
                    for (const neighbor of groups[curr].connectedTo) {
                        if (!groups[neighbor].merged) {
                            groups[neighbor].merged = true;
                            componentIndices.push(neighbor);
                            stack.push(neighbor);
                        }
                    }
                }
                // 5. Merge component
                // Sort by index to maintain some stability/sequential order
                componentIndices.sort((a, b) => a - b);
                const combinedAtoms: IAtom[] = [];
                const titles: string[] = [];
                for (const idx of componentIndices) {
                    const g = groups[idx];
                    combinedAtoms.push(...g.atoms);
                    // Use the first atom's info for the title part
                    titles.push(`${g.atoms[0].resn}:${g.atoms[0].resi}`);
                }
                const mergedTitle = titles.join("-");
                terminalNodes.push(
                    new TreeNode({
                        title: mergedTitle,
                        type: TreeNodeType.Compound,
                        model: combinedAtoms,
                        visible: true,
                        focused: false,
                        viewerDirty: true,
                        treeExpanded: false,
                        selected: SelectedType.False,
                    })
                );
            }
        }
        // Organize the flat list into a hierarchy
        let fileContents = organizeNodesIntoHierarchy(
            terminalNodes.toArray(),
            frameTitle
        );
        // Clean up and add to the list of frames
        fileContents = cleanUpFileContents(fileContents);

        fileContentsAllFrames.push(fileContents);
    }

    return Promise.resolve(fileContentsAllFrames);
}

/**
 * Given molecular data from the main thread, convert it into a TreeNode object
 * divided by component (protien, compound, solvent, etc.). For mol2 files.
 *
 * @param  {IMolData} data The molecular data.
 * @returns {Promise<TreeNode>} The divided molecule.
 */
async function divideMol2AtomsIntoDistinctComponents(
    data: IMolData
): Promise<TreeNodeList> {
    // For Mol2, just assume its a compound. Assign chain A to all atoms.
    const molName = getFileNameParts(data.fileInfo.name).basename;

    // Put it all in a compound
    // let compoundsByChain = organizeSelByChain({}, molWithAtomsToDivide, "Compound");

    const molNode = _getDefaultTreeNode(molName, NodesOrModel.Model);
    molNode.model = data.fileInfo;
    molNode.type = TreeNodeType.Compound;
    const rootNode = organizeNodesIntoHierarchy([molNode], molName);
    return new TreeNodeList([rootNode]);
}

/**
 * Clean up the treeNode in preparation for sending it back to the main
 * worker.
 *
 * @param  {TreeNode} treeNode The treeNode to clean up.
 * @returns {TreeNode} The cleaned up treeNode.
 */
function cleanUpFileContents(treeNode: TreeNode): TreeNode {
    if (treeNode.nodes) {
        // Iterate through organizedAtoms. If object and has no keys, remove it.
        // If list and has length 0, remove it.
        // mol_filter_ok
        treeNode.nodes = treeNode.nodes.filter((m: TreeNode) => {
            let totalSubItems = 0;
            totalSubItems += m.nodes ? m.nodes.length : 0;
            totalSubItems += m.model ? (m.model as IAtom[]).length : 0;
            return totalSubItems > 0;
        });

        // Clean up issues. If it's text has "undefined:", replace that with "".
        // If it has more than one node, add plural to text in some cases.
        treeNode.nodes?.forEach((m: TreeNode) => {
            if (
                m.nodes &&
                m.nodes.length > 0 &&
                ["Metal", "Ion", "Lipid", "Compound"].indexOf(m.title) > -1
            ) {
                m.title += "s";
            }

            // Replace "undefined:" in m.text with ""
            if (m.title) {
                m.title = m.title.replace(/undefined:/g, "");
            }
            m.title = m.title.replace(/undefined /g, "");
        });
    }

    if (COLLAPSE_ONE_NODE_LEVELS) {
        treeNode = collapseSingles(treeNode);
    }

    return treeNode;
}

/**
 * Adds the parent ids to the nodes.
 *
 * @param  {TreeNode} treeNode  The treeNode to add the ids to.
 */
function addParentIds(treeNode: TreeNode) {
    const allNodes = new TreeNodeList([treeNode]).flattened;
    allNodes.forEach((anyNode: TreeNode) => {
        if (anyNode.nodes && anyNode.nodes.length > 0) {
            anyNode.nodes.forEach((node: TreeNode) => {
                node.parentId = anyNode.id;
            });
        }
    });
}

waitForDataFromMainThread()
    .then(async (molData: IMolData[]) => {
        if (molData.length === 0) {
            sendResponseToMainThread([]);
            return;
        }

        // Format MUST be pdb or mol2 here. Should have converted using
        // openbabel first.
        for (const molDatum of molData) {
            if (["pdb", "mol2"].indexOf(molDatum.format) === -1) {
                throw new Error(
                    `Format must be pdb or mol2. Got ${molDatum.format}.`
                );
            }
        }

        // NOTE: Ligands already divided into frames. No need to divide them.

        const promises = molData.map((d) => {
            if (d.format === "pdb") {
                return dividePDBAtomsIntoDistinctComponents(d);
            }

            // TODO: Need to implement this for mol2 files.
            return divideMol2AtomsIntoDistinctComponents(d);
        });

        const organizedAtomsFramesList = await Promise.all(promises);

        // Add source to all nodes
        for (let i = 0; i < organizedAtomsFramesList.length; i++) {
            for (let j = 0; j < organizedAtomsFramesList[i].length; j++) {
                organizedAtomsFramesList[i].get(j).src =
                    molData[i].fileInfo.name;
            }
        }

        // Merge into one list
        const organizedAtomsFrames = organizedAtomsFramesList[0];
        for (let i = 1; i < organizedAtomsFramesList.length; i++) {
            organizedAtomsFrames.extend(organizedAtomsFramesList[i]);
        }

        let organizedAtomsFramesFixed = new TreeNodeList();
        organizedAtomsFrames.forEach((organizedAtoms: TreeNode) => {
            organizedAtoms.id = randomID();

            const nodesToConsider = new TreeNodeList([organizedAtoms]);
            if (organizedAtoms.nodes) {
                nodesToConsider.extend(organizedAtoms.nodes);
            }

            nodesToConsider.forEach((node) => {
                // TODO: In theory, you shouldn't need to set styles here,
                // because they get reset in the main thread based on the
                // current styles in the viewer (see addToMainTree). And yet
                // when I remove styles setting from the worker, the solvent no
                // longer appears in the viewer. I tried to figure out why, but
                // struggled to find a solution. So I'm leaving it here for now.
                // NOTE: I'm pretty sure styles now no longer needs to be set
                // here, but leaving because I hope to refactor later.
                switch (node.type) {
                    case TreeNodeType.Protein:
                        addMolTypeAndStyle(node, defaultProteinStyle);
                        break;
                    case TreeNodeType.Nucleic:
                        addMolTypeAndStyle(node, defaultNucleicStyle);
                        break;
                    case TreeNodeType.Compound:
                        addMolTypeAndStyle(node, defaultLigandsStyle);
                        break;
                    case TreeNodeType.Metal:
                        addMolTypeAndStyle(node, defaultMetalsStyle);
                        break;
                    case TreeNodeType.Lipid:
                        addMolTypeAndStyle(node, defaultLipidStyle);
                        break;
                    case TreeNodeType.Ions:
                        addMolTypeAndStyle(node, defaultIonsStyle);
                        break;
                    case TreeNodeType.Solvent:
                        addMolTypeAndStyle(node, defaultSolventStyle);
                        break;
                }
            });

            addParentIds(organizedAtoms);

            organizedAtomsFramesFixed.push(organizedAtoms);
        });

        organizedAtomsFramesFixed = organizedAtomsFramesFixed.filter(
            (o: TreeNode) => {
                const hasNodes = o.nodes !== undefined && o.nodes.length > 0;
                const hasModel =
                    o.model !== undefined && (o.model as IAtom[]).length > 0;
                return hasNodes || hasModel;
            }
        );

        organizedAtomsFramesFixed.terminals.forEach((node: TreeNode) => {
            // If node.model is a list, it's a list of atoms that need to be
            // converted to pdb.
            if (Array.isArray(node.model)) {
                // Convert to pdb
                node.model = convertIAtomsToIFileInfoPDB(node.model);
            } else {
                // It's mol2, already in text format.
                node.model = {
                    name: "tmp.mol2",
                    contents: (node.model as IFileInfo).contents,
                } as IFileInfo;
            }
        });

        // const JSZip = await dynamicImports.jsZip.module;
        // const zip = new JSZip();
        // organizedAtomsFramesFixed.terminals.forEach(
        //     (node: TreeNode, idx: number) => {
        //         if (node.model) {
        //             zip.file("tmp.zip", (node.model as IFileInfo).contents);
        //             // const content = zip.generate({type : "string"});

        //             // Generate the compressed string
        //             // eslint-disable-next-line promise/no-nesting
        //             const ttt = zip.generateAsync({ type: "base64" })
        //             .then(function (
        //                 compressedString: string
        //             ) {
        //                 // Use or store your compressedString somewhere
        //                 console.log(compressedString);
        //                 return;
        //             });
        //         }
        //     }
        // );

        sendResponseToMainThread(organizedAtomsFramesFixed.serialize());

        return;
    })
    .catch((err: any) => {
        throw err;
    });